jerclark
2016-03-15 01:05
has joined #general
bruno-c
2016-03-15 01:08
has joined #general
bruno-c
2016-03-15 01:11
yooo
bruno-c
2016-03-15 01:11
okay this is good i?m on slack like 24/7
nathaniel
2016-03-15 01:12
has joined #general
bruno-c
2016-03-15 01:13
some of the links i was showing earlier
nathaniel
2016-03-15 01:15
Thanks?I will take a look through these tomorrow
bruno-c
2016-03-15 01:16
so much stuff. there?s also a good post on piazza with links to a 100 other places
bruno-c
2016-03-15 01:16
a totally unrelated subject that i recently found interesting was spending habits
bruno-c
2016-03-15 01:16
what people buy and where. this one about canada is really interesting http://www.cbc.ca/radio/undertheinfluence
bruno-c
2016-03-15 01:17
> In the United States, New Hampshire consumes the most beer. Followed by North Dakota and Montana.
bruno-c
2016-03-15 01:17
> The state that consumes the most spirits is ? again - New Hampshire. Next comes Washington D.C. then Delaware.
bruno-c
2016-03-15 01:17
> Washington D.C. consumes the most wine. A very close second is Idaho. Bit of a dark horse there. And #3 coming around on the outside - is New Hampshire.
bruno-c
2016-03-15 01:18
> New York doesn't just buy a few more luxury watches, it buys 171% more luxury watches than the national average. No other American city comes close.
bruno-c
2016-03-15 01:19
not sure where they got their data but it highlights really interesting social trends, e.g that in new york that status symbol is an expensive watch. in boston it?s private schools, in sf it?s club memberships
bruno-c
2016-03-15 01:22
@bruno-c pinned a message to this channel.
jerclark
2016-03-15 02:03
Thanks for the manifest, Bruno?I?ll also peruse.
jerclark
2016-03-15 02:06
Check out this daily report viz about MBTA
nathaniel
2016-03-15 19:46
Created github repo for us: https://github.com/nathanielburbank/melodictransport
nathaniel
2016-03-17 19:57
I spent awhile looking through these today.
nathaniel
2016-03-17 19:58
If wanted to do something with song metadata, the columbia million song data looks promising.
nathaniel
2016-03-17 19:59
If we wanted to do something with playlist data (e.g., an analysis of road trip playlists) the Music APIs list has a lot of good resources, although we?d need to signup for and learn how to parse the API before we could do anything useful.
bruno-c
2016-03-18 01:55
'
jerclark
2016-03-19 23:18
I?ll check out the data - in the meantime, I?ve been trying to think of some interesting questions that we could try to answer about music and/or transportation.
jerclark
2016-03-19 23:19
The only questions I?ve come up with are actually about the relationship between housing costs and distance various points of interest.
jerclark
2016-03-19 23:24
Which has nothing to do with either music or transportation, I guess.
jerclark
2016-03-19 23:26
If we do song metadata, how about things like ?number of song titles containing certain words over time? - display trends about song titles, similar to that baby name thing we were shown early in the semester.
jerclark
2016-03-19 23:26
We could do similar things with other dimensions (song length, etc).
jerclark
2016-03-19 23:27
I?ll try to think of more questions that might be interesting to investigate.
bruno-c
2016-03-20 12:46
hmm that?d be pretty interesting!
bruno-c
2016-03-20 12:47
some kind of word cloud or graph that changes along a time axis
bruno-c
2016-03-20 14:12
> The MSD team is proud to partner with musiXmatch in order to bring you a large collection of song lyrics in bag-of-words format, for academic research. All of these lyrics are directly associated with MSD tracks: you can correlate them with all the data contained in the dataset such as similar artists, tags, years, audio features, etc.
bruno-c
2016-03-20 14:12
pretty cool stuff there
jerclark
2016-03-20 14:14
Lyrics would be a really rich data domain.
bruno-c
2016-03-20 14:14
for sure yeah. there?s a usage example here http://blog.last.fm/2011/06/22/lyric-clouds-genre-maps-and-distinctive-words
bruno-c
2016-03-20 14:17
ungh site is half dead
jerclark
2016-03-20 14:20
Did you do any more thinking about spending trends?
bruno-c
2016-03-20 14:25
i still think it would be really interesting data to explore . there is quite a bit of data available for the us
bruno-c
2016-03-20 14:26
not sure about the granularity of the detail though
bruno-c
2016-03-20 14:28
stuff like that :simple_smile:
bruno-c
2016-03-20 14:29
what?s interesting is that you have many dimensions to play with ? age, location, categories of things
bruno-c
2016-03-20 14:29
race
jerclark
2016-03-20 14:41
Any thoughts about exploring relationships between l urban housing costs and proximity to various point of interest? I'm sure that's been worked on, but it could lead to interesting visualizations.
bruno-c
2016-03-20 14:43
hmm, so rental / property values compared to say proximity to parks or schools or museums?
bruno-c
2016-03-20 14:43
are you shopping for a house Jeremy? :wink:
jerclark
2016-03-20 14:43
Transit stations, liquor stores, grocery stores, brothels, etc.
bruno-c
2016-03-20 14:44
lol oh yeah
jerclark
2016-03-20 14:45
Dude I'm anchored in place at the moment :)
jerclark
2016-03-20 14:48
@jerclark uploaded a file: https://melodictransport.slack.com/files/jerclark/F0U29SMFF/housing_cost_sketch.jpg and commented: This is a quick sketch...could be a choropleth of urban housing lots, shaded by cost. And user interaction could draw attention to certain points of interest
jerclark
2016-03-20 14:48
The thing in the middle is a zoom in of city streets on a map
bruno-c
2016-03-20 14:49
what question would it answer!
bruno-c
2016-03-20 14:49
totally using the cs171 our model slides now
jerclark
2016-03-20 14:52
Is there a correlation between proximity to a certain type of point of interest and housing cost? What is that correlation (higher cost/lower cost)?
bruno-c
2016-03-20 14:53
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F0U2A7HFZ/our-model.png
bruno-c
2016-03-20 14:53
@bruno-c pinned their Image https://melodictransport.slack.com/files/bruno-c/F0U2A7HFZ/our-model.png to this channel.
bruno-c
2016-03-20 14:53
i just remembered that thing
bruno-c
2016-03-20 14:53
yeah, that?d be a good question
jerclark
2016-03-20 14:54
Nice...yeah. I've only been thinking about questions at this point. I like the music domain, but haven't come up with any questions that seem anything more than novel. Which is probably OK, just trying to think of some other topics that have a different sociological posture.
bruno-c
2016-03-20 14:55
same here
bruno-c
2016-03-20 14:56
the course definitely has more of a data analytics type of slant, not so much a computational art kind of slant
bruno-c
2016-03-20 14:56
i can make music toys on my arduino and so on :simple_smile: so definitely not married to that
jerclark
2016-03-20 14:56
Trends in music over time could highlight some interesting stuff (like the word "peace" appearing during the 60's and 70s or something like that)
bruno-c
2016-03-20 14:57
yeah, or ?love? or ?dance"
bruno-c
2016-03-20 14:57
or nay nay
jerclark
2016-03-20 14:57
exactly!
jerclark
2016-03-20 14:58
But I agree with you about the focus of the class.
jerclark
2016-03-20 14:59
Have you ever played "electroplankton" on Nintendo DS? You might like it...
bruno-c
2016-03-20 15:24
i did! yeah that?s a really cool game
jerclark
2016-03-20 19:56
Thanks Bruno. So I'm assuming for tomorrow we need:
-questions to answer
-viable data set examples/apis
jerclark
2016-03-20 19:56
-rough sketches
jerclark
2016-03-20 19:56
-maybe a rough branching storyboard
bruno-c
2016-03-20 19:56
i think they simply ask for a couple paragraphs
bruno-c
2016-03-20 19:56
nothing so detailed
bruno-c
2016-03-20 19:57
a rough idea + what?s the dataset
jerclark
2016-03-20 19:57
Ok...that's good. Starts with a couple questions to answer, though - eh?
bruno-c
2016-03-20 19:57
that?s my understanding yeah
jerclark
2016-03-20 20:02
Alright...ill keep sending questions along
jerclark
2016-03-20 23:13
Boston housing value data - full list, good data.
jerclark
2016-03-20 23:15
Looked for something similar in Montreal - found this.
jerclark
2016-03-20 23:16
Other questions we might ask: What impact to property value does proximity to multiple points of interest have? (for example, does being close to a park and a transit stop have a cumulative effect?)
jerclark
2016-03-20 23:17
Could also break down property value by "lot" or "parcel" value (the value of an individual property) OR by price/area (square feet/meters).
nathaniel
2016-03-21 02:07
Sorry to be MIA guys ? was traveling this weekend.
nathaniel
2016-03-21 02:07
Lot?s of interesting questions here.
nathaniel
2016-03-21 02:08
This is the assignment we have for tomorrow: Please talk to your project team about your final project. Decide on a project title and write a 1-2 paragraph abstract. The abstract should include your motivation for doing this project, what you are trying to do and which goal you want to achieve. Also include a description about your planned datasets and where you will acquire them.
Once per team, please submit the project proposal by filling out this google form: http://goo.gl/forms/C0IeJWci15 (Links to an external site.) (Please only fill this out once per team, not for each team member individually!)
nathaniel
2016-03-21 02:12
Of the top of my head, Jeremy, I like your housing-costs vs. amenities idea. I?m sure it has been studied before, that does not make it a bad visualization.
nathaniel
2016-03-21 02:17
My other idea (going back to melodic transport) was to track the songs that are most often included ?road trip? playlists overtime.
nathaniel
2016-03-21 02:18
Once you had a song data set, you could also do some analysis the musical elements that make up traveling songs.
nathaniel
2016-03-21 02:18
However, I have not found a good data set of public shared playlists yet.
jerclark
2016-03-21 02:29
I definitely think the housing cost stuff has probably been studied/visualized?but I do think some creative analysis can be layered on top of it. But I partially mentioned it just because I thought there would be lots of data.
jerclark
2016-03-21 02:31
As for the road trip thing, there is this api which appears to have ?listening? data - though not necessarily organized by event like road tripping.
jerclark
2016-03-21 02:32
(From the digest of music API links that bruno sent)
jerclark
2016-03-21 02:34
In any case, I guess we need to come to some consensus - pick a topic we can at least all get behind; one with sufficient interesting data and possibilities for visualization.
jerclark
2016-03-21 02:35
Should we try to come to agreement by tomorrow afternoon and then circulate some draft proposals
bruno-c
2016-03-21 11:50
Had to do stuff yesterday. So, alright well. How about the customer expenditure stuff? Where how people spend their money? How does it vary from region to region, or across years?
bruno-c
2016-03-21 11:55
The real estate idea is good too but it almost seems like that's already being done on sites like trulia and Zillow
bruno-c
2016-03-21 11:55
Going back to the model, after we define questions we should explore available solutions
bruno-c
2016-03-21 11:57
At any rate I don't think they expect people to have all these things answered so we can always adjust as we go forward
jerclark
2016-03-21 22:54
I'm not married to the real estate thing. I think customer expenditures is interesting too...haven't looked at the data set yet.
jerclark
2016-03-21 22:56
Do you know of any data sets about that kind of thing?
bruno-c
2016-03-21 22:57
there?s quite a bit of data there
jerclark
2016-03-21 22:58
Was just browsing my way there...
jerclark
2016-03-21 22:58
Also just realizing you already posted that!
jerclark
2016-03-21 22:58
Do they have something similar for canada?
bruno-c
2016-03-21 22:58
yes i believe statistics canada has similar data
jerclark
2016-03-21 22:59
Should we think of another dimension that might be interesting to correlate?
bruno-c
2016-03-21 22:59
well the ones from BLS have a whole bunch of dimensions.. age.. income, education
jerclark
2016-03-21 22:59
Yeah - just seeing that now
bruno-c
2016-03-21 23:04
if we manage to get the data spread across some years, we could also theoretically highlight what happens when policies and laws change
bruno-c
2016-03-21 23:04
on an economic level
jerclark
2016-03-21 23:04
Good point - similar to the lyrics and baby names.
jerclark
2016-03-21 23:05
Did you have some questions in mind?
bruno-c
2016-03-21 23:06
How do various social factors such as age, race, education level affect consumer expenditure? What is the effect of changes in economic policies on consumer expenditure?
bruno-c
2016-03-21 23:06
etc
jerclark
2016-03-21 23:06
Seasons/Weather might be interesting as well...
bruno-c
2016-03-21 23:07
there has to be somewhat more profund insight to be found here but it?s not immediately obvious
bruno-c
2016-03-21 23:07
yeah, or location / region
jerclark
2016-03-21 23:07
Like what do people buy during specific season (or lack of seasons)....Also, would be interesting to see if things like insurance expenditures go up after big weather events
bruno-c
2016-03-21 23:07
that looks like the most complete dataset
bruno-c
2016-03-21 23:10
i think you can find out how much underwear is being bought!
bruno-c
2016-03-21 23:10
CLA-Clothing and sewing materials
Unique records in CLA are defined by the combination of NEWID, SEQNO, and ALCNO.
CLOTHYA is an item code variable for the type of clothing expenditure (suits, dresses,
undergarments, etc.)
CLOTHMOA is an indicator variable for the month in which the purchase occurred.
CLOTHXA is the amount of the clothing expense.
jerclark
2016-03-21 23:10
This looks rich.
jerclark
2016-03-21 23:10
I don't use stata or sas...do you?
bruno-c
2016-03-21 23:11
i don?t know any of those data types
jerclark
2016-03-21 23:11
Yeah...
jerclark
2016-03-21 23:11
I have a friend doing a Phd and she uses stata
bruno-c
2016-03-21 23:12
my first instinct would be to probably write programs to dump everything into a sqllite or postgres or mysql database
jerclark
2016-03-21 23:12
Is a statistics package of some sort.
bruno-c
2016-03-21 23:12
ah
nathaniel
2016-03-21 23:12
I?ve used Stata
jerclark
2016-03-21 23:12
Sure dumping to a DB makes sense.
jerclark
2016-03-21 23:12
Are you looking at the diary files?
nathaniel
2016-03-21 23:13
It?s not to too bad if you know how to use any of the major statistical packages
jerclark
2016-03-21 23:13
Not up to speed on any of em.
bruno-c
2016-03-21 23:13
yeah looks like there?s a zillion datapoints in there
bruno-c
2016-03-21 23:15
a lot more than i thought.
nathaniel
2016-03-21 23:22
Okay ? I?lm looking through the bls survey dump as well but I?m still not quite sure what I?m seeing. There?s a lot of data in here, but it?s hard to make sense of what it represents.
nathaniel
2016-03-21 23:23
I get that it?s survey data about expenditures, but is there any information about respondents? As far as I can tell it?s just ?American consumers?
bruno-c
2016-03-21 23:25
it?s american consumers, but with many, many dimensions
nathaniel
2016-03-21 23:26
Hmm..I think the table on Page 2 of this may do a good job of summarizing it. http://www.bls.gov/cex/csxann12.pdf
nathaniel
2016-03-21 23:27
What I?m seeing so for far is many dimensions of expenditures for ?average american consumers?. But what I?m not sure is if you can say, compare consumers in MA with consumers in NY.
nathaniel
2016-03-21 23:28
Which I think would be more interesting that pure aggregate changes over time.
jerclark
2016-03-21 23:44
I think that data is available: see p14 of this document: http://www.bls.gov/cex/2014/csxintvw.pdf
jerclark
2016-03-21 23:46
There is some serious data here...it'll take me a while to get my head around it.
nathaniel
2016-03-21 23:48
I think what?s that saying is that while, yes, they do identify state of the respondents you can?t make accurate state-by-state inferences because they do enough sampling in every state to get representative state-level stats.
nathaniel
2016-03-21 23:54
Here?s a totally different idea that I think would be easier to wrap our head?s around. Take a look at this: https://data.cityofboston.gov/Finance/Employee-Earnings-Report-2014/4swk-wcg8
nathaniel
2016-03-21 23:54
It looks like there?s comprehensive data from 2010-2015 for all city of Boston employees
jerclark
2016-03-21 23:55
You might be right
nathaniel
2016-03-21 23:55
You can see who earns the most overall, who earns the most from overtime, who?s salary is going up, going down, etc for 22,000 city employees
jerclark
2016-03-21 23:56
The city of boston data is really easy to parse.
jerclark
2016-03-21 23:56
It was the same for the housing price data
nathaniel
2016-03-21 23:58
It?s not a huge amount of data, but I can easily imagine 4 or visualizations, by employee type, by department, location etc
bruno-c
2016-03-21 23:58
¯\_(?)_/¯
nathaniel
2016-03-21 23:58
And there?s probably more that we could supplement it with from other sources.
bruno-c
2016-03-21 23:59
not feeling it :simple_smile: i could literally throw that in excel and start making charts
bruno-c
2016-03-22 00:00
too easy no?
nathaniel
2016-03-22 00:00
Perhaps
bruno-c
2016-03-22 00:00
actually i don?t have to, the visualization is built into the site
jerclark
2016-03-22 00:01
Fair point.
jerclark
2016-03-22 00:02
But - we could do things like cross reference their zipcodes and see how much they spend on housing.
jerclark
2016-03-22 00:02
We could also cross reference the property value data from the city of boston.
nathaniel
2016-03-22 00:03
Or to which neighborhoods the employees come from
nathaniel
2016-03-22 00:03
But ? I?m not wed to this idea at all
nathaniel
2016-03-22 00:03
Just trying to come up with a question that seems attainable
jerclark
2016-03-22 00:04
I see your point for sure, bruno - kinda simple, viz already available...But I'm also concerned about the time required to make sense of that CE data.
jerclark
2016-03-22 00:04
Were you able to make quick inroads into how it's setup?
bruno-c
2016-03-22 00:07
i?ve been reading about the different datapoints available in there. I have no clue regarding the detail of how exactly it?s stored
jerclark
2016-03-22 00:07
OK.
bruno-c
2016-03-22 00:07
i?m not particularly preoccupied by that (personally) but maybe that?s optimistic
nathaniel
2016-03-22 00:08
preoccupied?
bruno-c
2016-03-22 00:12
what i mean is i?m not concerned that i don?t know the details of the data format because we?ll figure out a way to get to it, essentially
nathaniel
2016-03-22 00:12
ok
nathaniel
2016-03-22 00:22
Right ? I?m sure we could figure it out. But you still need some sort of high level question that you?re trying to answer. I?m not seeing that currently with the BLS data?but if you got one I?m all ears.
jerclark
2016-03-22 00:23
Look back around 7:04 PM...
jerclark
2016-03-22 00:24
Some high-level questions....
bruno-c
2016-03-22 00:24
yeah those were very high level yes
jerclark
2016-03-22 00:25
Correlating events with spending habits could be really interesting (9/11, natural disasters, sports championships, etc...)
bruno-c
2016-03-22 00:26
yeah. i mean if we all think that the dataset is interesting, we can leave it at a high level question, or simply imply that?s it?s going to be an exploratory type of visualization
bruno-c
2016-03-22 00:26
where many questions and correlations can be discovered ? rather than a more story telling approach
nathaniel
2016-03-22 00:29
Can we jump on a google hangout for a moment?
bruno-c
2016-03-22 00:30
i have a really bad cold lol
bruno-c
2016-03-22 00:33
alright got my headphones
nathaniel
2016-03-22 00:33
That?s okay ? I just think it would be helpful to talk in person for just a bit. It does not need to be long
nathaniel
2016-03-22 00:33
(Creating google hangout ? may take a sec)
nathaniel
2016-03-22 00:36
Sorry ? having errors with google hang out on my end
bruno-c
2016-03-22 00:36
same here yeah
nathaniel
2016-03-22 00:39
Okay ? I don?t know. Perhaps this video call is not meant to be!
bruno-c
2016-03-22 00:39
trying to initiate one here
jerclark
2016-03-22 00:40
I can join in a few...gotta get tikes in bed
bruno-c
2016-03-22 00:54
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F0UCS15TN/wishy_washy_paragraph.txt
nathaniel
2016-03-22 01:00
That?s great
nathaniel
2016-03-22 01:00
I?ll expand on that a little bit
bruno-c
2016-03-22 01:00
:+1:
jerclark
2016-03-22 01:09
I think this is a good place to start, but I don't think we should disqualify the possibility of 'telling a story' - I think the focus on that aspect during the midterm suggests we should at least consider it.
nathaniel
2016-03-22 01:09
nathaniel
2016-03-22 01:14
Feel free to dive in and edit it if you want...
jerclark
2016-03-22 01:14
Sure.
jerclark
2016-03-22 01:15
It currently uses "Our goal" twice...just looking for another sentiment.
jerclark
2016-03-22 01:20
Should we include the possibility of visualizing spending data over time compared with various timelines of historical events (for example, a line chart with displaying spending data over time, with a user definable timeline underneath the x-axis displaying things like "party of current president", "stock market values", etc...)
nathaniel
2016-03-22 01:21
Sure ? I just could not quite figure out how to phrase that
jerclark
2016-03-22 01:21
Let me take a crack at it:
nathaniel
2016-03-22 01:21
Sure ? have at it
nathaniel
2016-03-22 01:22
I?m going to step away for a few min. Be back in 5.
jerclark
2016-03-22 01:31
Added a sentence...
nathaniel
2016-03-22 01:31
looks great
jerclark
2016-03-22 01:32
Cool...you think it's sufficient for submission?
nathaniel
2016-03-22 01:33
Yes ? I just want Bruno to confirm
jerclark
2016-03-22 01:34
Sure...
bruno-c
2016-03-22 01:34
Looks good to me!
nathaniel
2016-03-22 01:34
Great
nathaniel
2016-03-22 01:35
Our studio TF is Alain?correct?
bruno-c
2016-03-22 01:35
Yup
nathaniel
2016-03-22 01:35
Do you guys mind if I do the submission?
jerclark
2016-03-22 01:35
Nope...have at it!
nathaniel
2016-03-22 01:37
It?s submitted
jerclark
2016-03-22 01:38
Cool. Thanks.
bruno-c
2016-03-22 01:39
Great! Alright I'm outta here for the night. This should be fun. Catch you guys later!
nathaniel
2016-03-22 01:40
Great talk to you soon.
jerclark
2016-03-23 02:25
Hey Guys - just watched lecture and started the ?team expecations? document as spelled out in lecture. It?s one of the HW items. Please edit as you see fit - I also left the last two items empty to see if you guys wanted to take a crack at em:
jerclark
2016-03-23 02:25
@jerclark shared a file: https://melodictransport.slack.com/files/jerclark/F0UPHCPH6/cs171projectteamexpectations
nathaniel
2016-03-24 22:35
This is great
nathaniel
2016-03-24 22:37
I made a few edits, but overall, looks good to me.
nathaniel
2016-03-24 22:38
Since it sounds like we have several deliverables due on Monday evening, can I suggest that we do a meeting (google chat) at somepoint this weekend?
nathaniel
2016-03-24 22:38
I?m very available except for Saturday morning.
jerclark
2016-03-24 23:42
Absolutely - I'm in.
jerclark
2016-03-24 23:44
I'm pretty available as well...most of the weekend
bruno-c
2016-03-24 23:50
Yeah weekends are good for me too
nathaniel
2016-03-25 00:34
Can I suggest Saturday at 3 pm?
jerclark
2016-03-25 00:34
Sounds good to me.
nathaniel
2016-03-25 00:35
Great ? I need to go, but if that works for Bruno as well we can chat then.
bruno-c
2016-03-25 00:35
yup
bruno-c
2016-03-25 00:35
all good
jerclark
2016-03-25 00:35
Cool. See you guys then. You gonna send a google invite?
bruno-c
2016-03-25 00:36
Sure. Who?s the Scheduler?! :stuck_out_tongue:
nathaniel
2016-03-25 00:36
Yeah, I will send one around
nathaniel
2016-03-25 00:36
Yes, clearly me. :simple_smile:
nathaniel
2016-03-25 00:36
Ttyl
nathaniel
2016-03-26 19:05
nathaniel
2016-03-26 19:06
nathaniel
2016-03-26 19:14
jerclark
2016-03-26 20:08
Hey Guys - here?s a signed copy of the team expecations. If you all concur, sign and re-post in slack.
jerclark
2016-03-26 20:08
@jerclark uploaded a file: https://melodictransport.slack.com/files/jerclark/F0VMDUCCC/team_expectations.pdf and commented: Includes Jeremy?s signature only.
bruno-c
2016-03-26 20:09
alright i?m gonna print that out
nathaniel
2016-03-26 20:10
I?ll sign it after you bruno
bruno-c
2016-03-26 20:15
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F0VMH75ML/agreement-bruno-jeremy.pdf
bruno-c
2016-03-26 20:15
LOL 7 megs
bruno-c
2016-03-26 20:15
forgot to run it through my pdf shrinker
nathaniel
2016-03-26 20:15
This is like a game of pdf telephone
bruno-c
2016-03-26 20:16
i wonder how many times before it becomes completely illegible
nathaniel
2016-03-26 20:16
We?ll see ? I?m going to have to capture it with my cell phone?
nathaniel
2016-03-26 20:23
@nathaniel uploaded a file: https://melodictransport.slack.com/files/nathaniel/F0VMF1DHA/signed_cs_171_project_team_expectations.pdf and commented: Just barely legible...
nathaniel
2016-03-26 20:24
I?ll put this on github as well
jerclark
2016-03-26 20:24
Perfect. Playing librarian too?!
nathaniel
2016-03-26 20:24
Oh right ? I?m not supposed that
nathaniel
2016-03-26 20:24
I forgot!
bruno-c
2016-03-26 20:24
lo
bruno-c
2016-03-26 20:24
l
jerclark
2016-03-26 20:24
We?ll let it slip this time
bruno-c
2016-03-26 20:25
see i like that part but you know? we all have a little librarian inside of us
nathaniel
2016-03-26 20:46
FYI ? there?s a lot more discussion of the process book component here : http://www.cs171.org/2016/project/
bruno-c
2016-03-26 20:46
for possible future use
nathaniel
2016-03-26 20:50
Thanks for posting that Bruno. That looks like a useful explanation?
bruno-c
2016-03-26 20:51
Now, if only I knew R
jerclark
2016-03-26 20:51
exactly.
nathaniel
2016-03-26 20:51
R is its whole on world.
nathaniel
2016-03-26 20:51
But it?s not two bad.
nathaniel
2016-03-26 20:52
In our case I imagine we?ll only want to be using R/Stata or anything really to produce cvs files in the right format
bruno-c
2016-03-26 21:15
wow. super raw data
bruno-c
2016-03-26 21:15
literally an ftp site with ascii dumps of everything they have
bruno-c
2016-03-26 21:16
consumer expenditure is under cx/
bruno-c
2016-03-26 21:16
lots of other cool things though
jerclark
2016-03-26 21:17
I just started looking at ?ap? = average price. So every top level item is another whole data set?
bruno-c
2016-03-26 21:18
yeah!
bruno-c
2016-03-26 21:18
gas prices since like 1970 hehe
bruno-c
2016-03-26 21:19
APU000074712 1973 M10 0.402
bruno-c
2016-03-26 21:19
APU000074715 2011 M05 4.059
bruno-c
2016-03-26 21:31
a few questions
bruno-c
2016-03-26 21:31
bruno-c
2016-03-27 01:07
this high dimensional stuff is a jargon nightmare
bruno-c
2016-03-27 01:07
i think whatever poster we end up putting together should have a good glossary at the ?explain it like i?m 5? level
jerclark
2016-03-27 01:51
Yes - I?m wrapping my head around the high-dimensional stuff?although it does look like some of it could be very applicable to our project.
jerclark
2016-03-27 02:21
FYI: I renamed ?bruno-questions.md? to ?questions.md?, and figured we could all just work on that file. I made some adjustments and committed them?if that?s not cool with y?all just let me know.
bruno-c
2016-03-27 02:49
yeah that?s a good idea
bruno-c
2016-03-27 02:49
hey if any of you are still around, i?m trying to wrap up the lab
bruno-c
2016-03-27 02:50
the brush part ? in the instructions they say :
bruno-c
2016-03-27 02:50
> Update the x domain of the stacked area chart every time the brushed event gets triggered. And don?t forget to update the chart by calling:
bruno-c
2016-03-27 02:50
>? Reload the webpage in your browser. You should be able to choose a specific time period in the timeline (context) and the stacked area chart (focus) will automatically zoom in and will show the details for the selected years.
bruno-c
2016-03-27 02:51
i don?t quite get how that?s all going to happen by just changing the domain. the axis, etc are bound to the domain but they?re not going to magically refresh themselves
bruno-c
2016-03-27 02:52
so my brushed event handler looks like
bruno-c
2016-03-27 02:52
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F0VLXR24B/-.js
bruno-c
2016-03-27 19:09
well that?s interesting. it does work.
bruno-c
2016-03-27 19:10
updateVis refreshes everything needed. well that?s great.
bruno-c
2016-03-27 20:08
unless there are objections, i?m going to check in some of the documents i?m finding during research into our repo
bruno-c
2016-03-27 20:08
it may or may not be useful later but might as well keep them all in one place
jerclark
2016-03-27 23:03
Sounds good.
jerclark
2016-03-27 23:03
Bruno, FYI all commit notifications should post automatically to the git slack channel...
bruno-c
2016-03-27 23:07
Sweet
nathaniel
2016-03-28 01:14
Can you guys see if you have access to this database with your Harvard logins? http://harvard.policymap.com.ezp-prod1.hul.harvard.edu/maps
nathaniel
2016-03-28 01:33
On the hangout: https://hangouts.google.com/call/d6vajph4qncljdgxdgfaqxt4lue
nathaniel
2016-03-29 00:54
bruno-c
2016-03-29 00:55
cool. i am writing my stuff in markdown right now
nathaniel
2016-03-29 00:55
Sure ? just posting as an FYI
bruno-c
2016-03-29 00:55
i?m gonna go paste it in there shortly.
jerclark
2016-03-29 01:09
Working on last sketch currently.
nathaniel
2016-03-29 01:11
Dataset section looks great
bruno-c
2016-03-29 01:12
thanks. i don?t think we need to go into more detail but you guys can let me know what you think
nathaniel
2016-03-29 01:12
I think that?s more than good enough for now.
bruno-c
2016-03-29 01:34
so other than this, can i be of assistance for the rest of the stuff? screencast, etc? i got some other homework to deal with as well and i turn into a pumpkin pretty early at night
nathaniel
2016-03-29 01:36
I think we?ve got everything we need except for the screencast, which Jeremy is working on? (or working on the sketches for.)
jerclark
2016-03-29 01:37
Hey Bruno - it looks like what you added is the ?individual? homework.
jerclark
2016-03-29 01:37
Is that what you intended?
bruno-c
2016-03-29 01:37
to the repo? yeah
jerclark
2016-03-29 01:37
OK.
bruno-c
2016-03-29 01:37
i wanted it in there because it?ll be useful for the poster thing
jerclark
2016-03-29 01:38
Sorry - just looked at the dataset thing in the google doc.
jerclark
2016-03-29 01:38
nvmd
jerclark
2016-03-29 01:38
Sorry about delay on the sketch?trying to think of something creative.
jerclark
2016-03-29 01:38
For the screencast, do we need to make slides?
nathaniel
2016-03-29 01:39
I made a basic framework
jerclark
2016-03-29 01:39
OK cool.
nathaniel
2016-03-29 01:39
(Feel free to edit or change it..
jerclark
2016-03-29 01:40
You guys saw sketch #1.
jerclark
2016-03-29 01:40
Here?s #2
jerclark
2016-03-29 01:41
@jerclark uploaded a file: https://melodictransport.slack.com/files/jerclark/F0W38838Q/photo_on_3-28-16_at_9.40_pm.jpg
nathaniel
2016-03-29 01:41
Looks great!
jerclark
2016-03-29 01:41
My handwriting + backwards projection makes it hard to read.
nathaniel
2016-03-29 01:42
I need to step away for few minutes. Will be back by 10pm?
jerclark
2016-03-29 01:42
Sure...
bruno-c
2016-03-29 01:43
looks like something you can see the future in
jerclark
2016-03-29 01:43
lol. I?ll keep cranking and try to bust out a screencast.
bruno-c
2016-03-29 01:43
really nice :+1:
bruno-c
2016-03-29 01:43
i love those radar charts
jerclark
2016-03-29 02:49
Nathaniel - I added the visualizations to the project plan writeup
jerclark
2016-03-29 02:49
And I?m uploading the screencast now.
jerclark
2016-03-29 02:49
Will circulate link when ready.
nathaniel
2016-03-29 02:51
It all looks great?
nathaniel
2016-03-29 02:52
(or should have said, the slides look great.)
jerclark
2016-03-29 02:59
I think we have everything to submit...
nathaniel
2016-03-29 02:59
Screen cast is great! Thanks for recording that...
jerclark
2016-03-29 02:59
No problem...
jerclark
2016-03-29 03:00
Can you submit the materials?
nathaniel
2016-03-29 03:00
Sure
nathaniel
2016-03-29 03:00
And I think we?re supposed to post a link to the screen cast on Piazza as well? (I?ll double check)
jerclark
2016-03-29 03:01
Thx.
jerclark
2016-03-29 03:01
Let me know if you need anything else from me...
nathaniel
2016-03-29 03:02
Can you put image files of the sketches in the Github directory?
jerclark
2016-03-29 03:02
Sure.
nathaniel
2016-03-29 03:02
Oh, I see
nathaniel
2016-03-29 03:03
you put them in the plan doc as well
nathaniel
2016-03-29 03:03
that?s fine
jerclark
2016-03-29 03:11
Added sketches and screencast to git anyhow.
nathaniel
2016-03-29 03:14
Yup I saw that.
nathaniel
2016-03-29 03:14
Project plan is submitted
jerclark
2016-03-29 03:18
Awesome...
jerclark
2016-03-29 03:18
Did you submit the ?team expectations? and the screencast too?
nathaniel
2016-03-29 03:18
Yes
jerclark
2016-03-29 03:19
Sweet. Thanks! Heading off to bed now...
nathaniel
2016-03-29 03:19
I put it all together in a single zip file
nathaniel
2016-03-29 03:19
Talk to you later?
jerclark
2016-03-29 03:19
Talk soon. Maybe tomorrow?
nathaniel
2016-03-29 03:19
Yeah, I guess.
jerclark
2016-03-29 03:19
K
jerclark
2016-03-29 03:19
Gnite for now!
jerclark
2016-03-30 02:11
Hey Guys - I just watched the lecture?Attached are my notes and some slide captures if you?re interested. NOTE: According to the last slide, the DCE poster is due on FRIDAY 11:59 PM.
jerclark
2016-03-30 02:12
@jerclark uploaded a file: https://melodictransport.slack.com/files/jerclark/F0WDMLH25/week_10_-_lecture_notes__poster_.pdf
jerclark
2016-03-30 02:12
We should probably meet tomorrow night to discuss.
nathaniel
2016-03-30 11:53
Thanks Jeremy ? this is helpful
nathaniel
2016-03-30 20:57
Tonight at 9 or 9:30 pm would work for me to discuss the poster. I could also do tomorrow evening at about the same time.
jerclark
2016-03-30 23:26
I can do 9:30 tonight or tomorrow
bruno-c
2016-03-31 00:28
can do either.
nathaniel
2016-03-31 01:24
Sorry for the late notice ? but something came up tonight that I need to deal with. So, if it?s the same to both of you, let?s do this tomorow night at 9:30
nathaniel
2016-03-31 01:24
Thanks
jerclark
2016-03-31 01:59
Perfect...just got off a call myself so I was delayed.
jerclark
2016-03-31 02:00
Does anybody want to create a 'template' poster (wireframe layout, basically)? I can take a crack at it, unless one of you think you have a good layout eye.
jerclark
2016-03-31 02:01
Is there a tool you think we should use to create this thing? (Probably something a little more "design savvy" than word or google docs)?
bruno-c
2016-03-31 03:54
i was kinda thinking that a web page might work too. you can basically save it as a very large image
bruno-c
2016-03-31 03:55
otherwise you?re getting into graphic design type tooling i guess, photoshop, etc
nathaniel
2016-03-31 15:31
Yeah?all of larger design stuff I?ve ever done has involved specialized software which I?d rather not getinto for this project.
nathaniel
2016-03-31 15:31
Doing it as a webpage could be fine
nathaniel
2016-03-31 15:32
Or we could just use google docs?I don?t know how good their layout engine is
nathaniel
2016-03-31 15:32
Anyway, I will talk to you both at 9:30 pm tonight
jerclark
2016-04-01 00:38
Sounds good.
jerclark
2016-04-01 00:42
I have and use indeisgn, but maybe a web layout is best.
jerclark
2016-04-01 01:24
My midterm project was somewhat ?poster-like? - maybe we could use something like this as the basic layout: https://labs.vocareum.com/web/11744/4439/LatestSubmission/Project/index.html
nathaniel
2016-04-01 01:28
Here?s a google hangout link for whenever you guys are ready https://hangouts.google.com/call/452vifxckff6dnmpyszhlc2o7qe
nathaniel
2016-04-01 01:35
We could also use google docs, with a modified presentation
nathaniel
2016-04-01 01:35
bruno-c
2016-04-01 01:36
hey guise
bruno-c
2016-04-01 01:37
apologies for the lateness
bruno-c
2016-04-01 01:37
sweet title nathaniel ?biting back? :simple_smile:
bruno-c
2016-04-01 01:37
i had something like that and then i was like? i dunno, do i wanna be clever about malaria? hehe
bruno-c
2016-04-01 01:38
love the concentric circles too
nathaniel
2016-04-01 02:14
https://qph.is.quoracdn.net/main-qimg- 288521e3f8a2a8ba0aaabf5d90cf33b2?convert_to_webp=true
bruno-c
2016-04-01 02:16
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F0X5204EN/pasted_image_at_2016_03_31_10_16_pm.png
bruno-c
2016-04-01 02:23
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F0X4Q9GKS/spider_chart2.jpg
nathaniel
2016-04-01 02:26
@nathaniel uploaded a file: https://melodictransport.slack.com/files/nathaniel/F0X4S4TJB/pasted_image_at_2016_03_31_10_26_pm.png
nathaniel
2016-04-01 02:35
nathaniel
2016-04-01 02:44
@nathaniel uploaded a file: https://melodictransport.slack.com/files/nathaniel/F0X508WM7/pasted_image_at_2016_03_31_10_44_pm.png
jerclark
2016-04-01 03:05
bruno-c
2016-04-01 03:20
alvaro valiño
bruno-c
2016-04-01 05:34
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F0X5E5ZS4/pasted_image_at_2016_04_01_01_34_am.png
jerclark
2016-04-01 13:27
Awesome! Looks like a little dog walking and 45 min without us did you some good!
nathaniel
2016-04-01 14:03
Really does look great! Thanks Bruno for all of the hard work?
bruno-c
2016-04-01 14:04
Np! I tweaked it a bit more after that but I think the layout works well
nathaniel
2016-04-01 14:06
Nope, I?m looking at the google doc now. It really does look good.
nathaniel
2016-04-01 14:06
I will go through and finalize the copy, but I will not change the layout
nathaniel
2016-04-01 23:29
Not sure what "La la la la la la hamburger necklace?? was supposed toe a placeholder for?
nathaniel
2016-04-01 23:29
supposed to be
jerclark
2016-04-01 23:29
I cracked up when I read that?I do think it?s a place holder.
bruno-c
2016-04-01 23:30
You got it. For, well, whatever the other ones have
bruno-c
2016-04-01 23:30
Like a one liner
nathaniel
2016-04-01 23:30
You mean like a sup title for the third visualization on the lower right?
bruno-c
2016-04-01 23:32
Yeah I figured a link and a title and short description
bruno-c
2016-04-01 23:32
I remember one of you saying you liked that one
nathaniel
2016-04-01 23:33
Jeremy ? if you can add a sub title for the MBTA visualization I think we?re good.
jerclark
2016-04-01 23:33
How about ?Parallel coordinates and colors represent time, location, line, and schedule in a single plot"
jerclark
2016-04-01 23:34
I?ll add to the document.
jerclark
2016-04-01 23:34
I see you?re on it.
nathaniel
2016-04-01 23:34
I have to take off though
nathaniel
2016-04-01 23:34
Feel free to edit or make any other changes.
jerclark
2016-04-01 23:35
Cool. I?ll make a small tweak. Thanks nathaniel!
nathaniel
2016-04-01 23:35
Thank you! Talk to you both tomorrow.
jerclark
2016-04-01 23:37
Can we say ?Creative Visualizations? instead of "original visualizations? at the bottom?
nathaniel
2016-04-01 23:37
Fine by me.
nathaniel
2016-04-01 23:37
Leaving now ? ttyl
jerclark
2016-04-01 23:37
Have fun!
jerclark
2016-04-01 23:49
Made a couple of tweaks and submitted: https://canvas.harvard.edu/courses/9586/assignments/62593
nathaniel
2016-04-02 03:38
Great ? thanks!
nathaniel
2016-04-02 17:01
Please submit your completed project plan (one submission per team is sufficient). Make sure to include the following:
2nd iteration on sketches, including new visualizations you might have added after the poster session
description of team roles
feature list
story board for interaction
webpage layout/storytelling
project timeline (with milestones when you are planning to finish which feature)
Additionally, you should also include the following:
Acquire your complete dataset, and add a description/screenshot/etc. to the project plan. You should have the dataset by the time of submission of this homework!
Please submit Photos of your Vis Exploration Poster Session Questionnaire (one for each team member) .
nathaniel
2016-04-02 17:23
nathaniel
2016-04-02 17:24
(We?re in the google hangout)
nathaniel
2016-04-02 17:26
Jeremy?
nathaniel
2016-04-02 17:31
Just to fill you in Jeremy? We created a task list (https://docs.google.com/document/d/1B_kCWLmDyJy7o2GSschADxjJ4skETx2MmbSv6lwnxqY/edit) for our project revisions and are planning on meeting again tomorrow evening to review.
nathaniel
2016-04-02 17:33
We?re planning on meeting again tomorrow night at 9:00 pm. (I?ll send out a calendar invite in a moment.)
jerclark
2016-04-02 20:37
Sounds good?did I miss a scheduled meeting? I?m sorry about that...
jerclark
2016-04-02 20:37
We said saturday at 3 didn?t we!
jerclark
2016-04-02 20:38
Ready for negative feedback?
jerclark
2016-04-02 20:40
Saturday at 1PM...
jerclark
2016-04-02 20:40
I got it in my calendar now.
jerclark
2016-04-02 20:44
Thanks for putting together the task list?I assigned the ?open items? to myself in the google doc. I?ll get working on the visualizations and story board.
jerclark
2016-04-02 20:44
Also - 9PM tomorrow is good for me.
nathaniel
2016-04-02 23:43
Sorry if I was not clearer about the time. Talk to you tomorrow.
jerclark
2016-04-03 18:11
Definitely my fault.
jerclark
2016-04-03 18:12
Anyhow - I?ve been thinking about a ?tree-based? representation instead of the house representation for a ?novel? look at spending. Here?s the data I think we?d need:
jerclark
2016-04-03 18:43
Category names, Subcategory names, avg subcategory item price by year, avg household income by year, avg subcategory item spend by year
bruno-c
2016-04-03 18:43
ah! i got something for you then
jerclark
2016-04-03 18:44
Bruno, if you?re online do you have 5 min to chat?
bruno-c
2016-04-03 18:46
sure, at 3pm ok?
jerclark
2016-04-03 18:46
yep
bruno-c
2016-04-03 18:46
pull from there ? if you run the website that is inside the dataset folder (using webstorm or whatever)
bruno-c
2016-04-03 18:47
you can see what i?ve been messing with this afternoon
jerclark
2016-04-03 18:47
That is awesome.
jerclark
2016-04-03 18:47
Nice work with the inflation thingy.
jerclark
2016-04-03 18:47
Let me have a look.
bruno-c
2016-04-03 18:48
in the db folder, there?s actually a full import of the data into a sqllite database. it?s totally raw i haven?t had the chance to add indexes or whatever
jerclark
2016-04-03 18:48
Would still like to chat at 3 if you have sec.
bruno-c
2016-04-03 18:48
for sure
jerclark
2016-04-03 18:48
cool.
bruno-c
2016-04-03 19:02
alright
nathaniel
2016-04-04 01:02
In the hangout
bruno-c
2016-04-04 01:03
Be right there finishing a couple bites
nathaniel
2016-04-04 01:55
bruno-c
2016-04-04 03:55
alright data explorer has been updated. if you select all the pieces, it should show available series. next step is actually showing the data, probably going to be a horizontal bar chart in the 4th colum
bruno-c
2016-04-04 03:55
not sure i can load the entire dataset upfront but if not i?ll figure out ways to break it up
bruno-c
2016-04-04 03:57
still glitchy but possible to play with it. see you tomorrow
jerclark
2016-04-05 01:02
Hey Gents
nathaniel
2016-04-05 01:02
Hi
bruno-c
2016-04-05 01:02
yo
nathaniel
2016-04-05 01:03
Sorry ? lost track of time. Let me a get a goto meeting url
nathaniel
2016-04-05 01:03
I mean google hangout
jerclark
2016-04-05 01:03
Sure.
bruno-c
2016-04-05 01:03
goto meeting?:)
nathaniel
2016-04-05 01:03
That?s what we use at work
bruno-c
2016-04-05 01:04
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F0XV09YJC/gpg3k75.jpg
jerclark
2016-04-05 01:05
jerclark
2016-04-05 01:07
@jerclark shared a file: https://melodictransport.slack.com/files/jerclark/F0XUVEM46/interaction-storyboard-1
bruno-c
2016-04-05 01:46
nah the questionnaire thing is one sheet per team member
bruno-c
2016-04-05 01:49
okay so the project plan says ?add a description / screenshot? for the dataset
nathaniel
2016-04-05 01:49
We?ll include a screenshot of your data explorer
bruno-c
2016-04-05 01:49
yeah that should totally work
bruno-c
2016-04-05 01:50
so hmm, i?ll send the ?photo? of my vis exploration poster session questionnaire, whatever tf that means
bruno-c
2016-04-05 01:51
we?re group 2 right?
nathaniel
2016-04-05 01:51
Yes
bruno-c
2016-04-05 01:52
i found a giant file with all the posters in it, i think
bruno-c
2016-04-05 01:53
this poster thing is so dumb. argh
nathaniel
2016-04-05 01:56
totally agree ? that link is helpful
nathaniel
2016-04-05 01:58
In terms of the photo thing, I plan on printing it out and doing a selfie with the printed doc
nathaniel
2016-04-05 01:58
:simple_smile:
bruno-c
2016-04-05 01:59
lol. i?m just annotating the pdf
nathaniel
2016-04-05 02:00
Oh ? there?s a video of the poster session
nathaniel
2016-04-05 02:03
(And, it?s clear that the photo part is just supposed to be a photo of your poster, not the questionnaire)
bruno-c
2016-04-05 02:12
and yet there?s a whole bunch of questionnaires in the zipped package
bruno-c
2016-04-05 02:12
so that?s handy
bruno-c
2016-04-05 02:18
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F0XVCKQJ3/poster-session-bruno-c.pdf
bruno-c
2016-04-05 02:20
there are some really nice posters from the dce
bruno-c
2016-04-05 02:20
fuuuuuu that must?ve taken forever
bruno-c
2016-04-05 02:21
that shape of stories one, i?m so jealous of their idea
nathaniel
2016-04-05 03:02
Here is my version of the feature list
nathaniel
2016-04-05 03:02
nathaniel
2016-04-05 03:05
Jeremy ? feel free to edit/change as you see fit, or we can submit as is.
jerclark
2016-04-05 03:08
I made some adjustments?.
jerclark
2016-04-05 03:08
Is it OK?
jerclark
2016-04-05 03:11
We were supposed to offer some Layout/Storytelling?should we not worry too much about that tonight?
jerclark
2016-04-05 03:13
Where is the poster gallery?
jerclark
2016-04-05 03:14
Forget it - I see earlier link.
nathaniel
2016-04-05 03:31
Your edits look fine.
nathaniel
2016-04-05 03:31
I think we?re good for now.
nathaniel
2016-04-05 03:31
Once you get your poster feedback done we should submit
jerclark
2016-04-05 20:44
I submitted a really basic layout/story document with the assignment (if you can call it that)
nathaniel
2016-04-05 20:49
Thanks for doing that ? I?m sorry I had neglected to notice that part of the assignment until the last moment.
jerclark
2016-04-08 15:33
Hey guys...I'm in DC visiting my brother. I can arrange to meet tomorrow, and you want to meet at another time, just give me a little advance notice. Looks like "redesign" is an innovation phase...
nathaniel
2016-04-08 15:34
Tomorrow in the late afternoon/early evening would actually work better for me than 1 pm
nathaniel
2016-04-08 15:34
Or early morning (8 or 9 am)
nathaniel
2016-04-08 21:26
Hi Guys ? I finally think I understand our tasks for the ?evaluation? compent of this project
jerclark
2016-04-08 21:27
I need to finish watching the lecture...I think they do some live examples.
nathaniel
2016-04-08 21:27
However, we need to have two separate meetings, one where Group 1 evaluates us, and another where we evaluate Group 3. Both sessions should be 30 mins long, and need to take place before 5 pm EST on Sunday
nathaniel
2016-04-08 21:28
I?m going to suggest times on Sunday morning, between 10 am and 1 pm; do either of you have any conflicts/objections to those times?
nathaniel
2016-04-08 21:28
(Everything is done by google hangouts)
jerclark
2016-04-08 22:04
That's fine with me....
jerclark
2016-04-08 22:29
Also, since I'm down here I took Monday off work to do the bulk of anything I need to deliver.
nathaniel
2016-04-09 00:00
Okay ? I?m flying back to DC on Monday evening?
bruno-c
2016-04-09 00:06
All good for me as well. Sunday should be fine.
nathaniel
2016-04-09 00:07
Okay ? so just to clarify; I?m going to cancel our meeting tomorrow @ 1pm I?m going to schedule several sessions on Sunday morning for us to work with other groups
nathaniel
2016-04-09 00:07
And then for us to talk about redesigns
nathaniel
2016-04-09 00:07
I will send out google calendar invites once confirmed
nathaniel
2016-04-09 01:31
I have proposed feedback meetings for 11 am and 11:30 on Sunday morning
nathaniel
2016-04-09 01:31
Will send out invites once confimed
jerclark
2016-04-09 02:01
Do you guys have a handle on heuristic evaluation?
nathaniel
2016-04-09 02:07
Ehh..kind of.
nathaniel
2016-04-09 02:08
Sorry for all the calendar emails; as it stands currently we?re evaluating group 3 at 11 am on Sunday.
nathaniel
2016-04-09 02:08
then we?re having group 1 evaluate us at 11:30 am.
nathaniel
2016-04-09 02:09
Then we can discuss among ourselves what to do with that feedback at noon.
jerclark
2016-04-09 02:29
OK.
jerclark
2016-04-09 02:30
I need to pick up my wife from the train station on sunday around 12:15-12:30...so I'll do my best to participate at noon. Frankly, this scheduling timeline is a little crazy.
jerclark
2016-04-09 02:36
Some heuristic links:
jerclark
2016-04-09 02:36
https://www.nngroup.com/articles/ten-usability-heuristics/ - this seems to be where the "category" of interaction was taken from in the example given in piazza.
jerclark
2016-04-09 02:45
This one seems to set the 'framework' of a heuristic evaluation:
nathaniel
2016-04-09 02:58
We can push back the 12 pm meeting; I was just trying to make it all in one block
nathaniel
2016-04-10 01:42
Just to give a schedule update ? we were not able to make the meeting with the group that is reviewing us work
nathaniel
2016-04-10 01:43
So I just sent them a copy of our project plan
nathaniel
2016-04-10 01:43
and they are going to review it offline and send us written feedback tomrrow
nathaniel
2016-04-10 01:44
So we still have our review session at 11 am, and then we can talk as a group as a bit.
nathaniel
2016-04-10 01:44
afterwords at 11:30, but that?s it for now.
jerclark
2016-04-10 02:34
Thanks for setting up Nathaniel...
bruno-c
2016-04-10 14:32
@bruno-c pinned @jerclark's GDocs Document https://melodictransport.slack.com/files/jerclark/F0XUVEM46/interaction-storyboard-1 to this channel.
jerclark
2016-04-10 14:32
Dude, group 3's stuff looks really good!
jerclark
2016-04-10 14:32
Just glanced at it so far...
bruno-c
2016-04-10 14:32
agreed
bruno-c
2016-04-10 14:33
where?s our file like that? i can?t find the link anymore :confused:
bruno-c
2016-04-10 14:37
be back in a bit
jerclark
2016-04-10 14:46
Think the closest thing we have is: https://github.com/nathanielburbank/melodictransport/tree/master/project-plan%20v2
jerclark
2016-04-10 14:46
The images are in separate files
jerclark
2016-04-10 14:50
Are they supposed to give us questions?
jerclark
2016-04-10 14:53
They have chosen a VERY basic data set.
jerclark
2016-04-10 14:54
I see they gave us scenarios...
jerclark
2016-04-10 14:54
I think the basic data set could be an advantage...
bruno-c
2016-04-10 14:54
so, i?m not sure i understood nathaniel?s update ? are they showing up in 5 minutes or is that the meeting that we?re not really having with them?
jerclark
2016-04-10 14:55
I think they're showing up in 5 minutes...its the 11:30 where group 1 was evaluating us that got cancelled.
bruno-c
2016-04-10 14:55
ok but the meeting in 5 minutes is for us, to give us feedback on their project, right?
jerclark
2016-04-10 14:55
Yes
bruno-c
2016-04-10 14:55
right
jerclark
2016-04-10 14:56
In the file that we're looking at now.
bruno-c
2016-04-10 14:56
ok. i?ll start annotating in there
jerclark
2016-04-10 14:56
We're supposed to be 'independently' evaluating, so I made a section in the document for each of us.
jerclark
2016-04-10 14:57
And I think we're supposed to evaluate the tasks they gave us.
nathaniel
2016-04-10 14:59
That?s correct
nathaniel
2016-04-10 14:59
11:30 meeting got canceled
nathaniel
2016-04-10 15:25
Jeramy: if you want to keep talking now we can
nathaniel
2016-04-10 15:25
or we can do it later
jerclark
2016-04-10 15:25
Sure...
jerclark
2016-04-10 15:25
I have a few minutes now...
nathaniel
2016-04-10 15:25
my afternoon is open and free
2016-04-10 15:25
@bruno-c has started a Google+ Hangout for this channel. https://plus.google.com/hangouts/_/yvq3zsbsenb2la24fohpacidpae.
nathaniel
2016-04-10 15:32
For a period of time, what historical events influenced the changes in spending?
Find a source of spending where that has increased/became more burdensome overtime?
jerclark
2016-04-10 15:34
Find a source of spending that has become more burdensome over time?
bruno-c
2016-04-10 15:34
:+1:
jerclark
2016-04-10 15:34
This can be express either as:
jerclark
2016-04-10 15:34
1) Needing to spend more for the same amount of goods/services, OR
bruno-c
2016-04-10 15:34
@bruno-c set the channel purpose: Explore changes in the price of bananas
jerclark
2016-04-10 15:35
2) Needing to spend the same amount for less goods/services.
jerclark
2016-04-10 15:35
"Amount" == "percentage of income", NOT "absoulte cost"
nathaniel
2016-04-10 16:16
nathaniel
2016-04-10 16:16
We just got the feedback from group 1
jerclark
2016-04-10 17:33
Thanks, NB. BTW, I added my evaluation to the doc that group 3 shared. Perhaps you guys can do the same when you're done, then we can tell group 2 and then print out that document for submission?
jerclark
2016-04-10 17:37
After reading the Group 6's report, perhaps we should make a separate document for ourselves to clean up?
jerclark
2016-04-10 17:39
In reading the assignment, I think just adding the results to their google doc is probably OK.
bruno-c
2016-04-10 19:00
looks like loading the entire dataset, with almost 400000 items, only takes a few seconds on my machine
bruno-c
2016-04-10 19:00
so for now that?s the strategy i?m gonna use until we feel it doesn?t make sense performance wise
bruno-c
2016-04-10 19:28
alright folks, we have some data. the dataset explorer ?works? now. you can drill down all the way to yearly data and see it ordered by year
bruno-c
2016-04-10 19:33
wow, white people drink a lot
bruno-c
2016-04-10 19:33
heh
bruno-c
2016-04-10 19:35
hmm still some glitches but at least we can play with the data a bit. i?ll keep cleaning that up
bruno-c
2016-04-10 22:34
k us presidents / parties and us natural disasters have been pulled more or less legally from the web :wink:
bruno-c
2016-04-10 22:34
dumped to json under datasets/events
jerclark
2016-04-10 22:38
Nice work!
jerclark
2016-04-10 22:38
Thx
jerclark
2016-04-10 22:52
As for redesign, the small multiples instead of the force directed tree is an interesting idea.
bruno-c
2016-04-11 01:04
soooo
nathaniel
2016-04-11 01:04
Hey
nathaniel
2016-04-11 01:06
Alain want?s us to bring back the house visualization
bruno-c
2016-04-11 01:06
oh no!
bruno-c
2016-04-11 01:06
lol
bruno-c
2016-04-11 01:06
cool idea but i don?t know how practical it is!
nathaniel
2016-04-11 01:07
It *is* novel, but I don?t see how it works.
bruno-c
2016-04-11 01:07
how did that happen anyways? he just emailed you?
nathaniel
2016-04-11 01:07
Also, it?s not like physical size corresponds well to cost
nathaniel
2016-04-11 01:07
I think you?re on the email as well
nathaniel
2016-04-11 01:07
Bruno+Jeremy+Nathan,
Good job on your full project plan.
Here are my comments:
- Good job on providing the team roles
- I am still not very clear on how the important historical events will be conveyed against the stacked area chart. Can you elaborate or give a gist as to how the historical events will be visually linked to the views?
- I would have liked to see a write-up on the views that you are planning to build. In addition, did you decide to abandon the house visualization? I felt that it was novel and that it had great potential
- I would have liked you to hone in more on your dataset in your write-up. For example, what metadata are you capturing? What data links are you creating?
Given your project plan, I feel that there are components that will help you in solidifying your direction:
1) Identify your primary and secondary tasks. What are the main goals of the visualization?
2) What story are you telling? How will the story be told? What text and images will you be using?
3) What is the goal of each view? What task(s) is it supporting?
As always, please let me know if you have any questions.
bruno-c
2016-04-11 01:09
yup, see it now.
jerclark
2016-04-11 01:09
Hey Guys...
jerclark
2016-04-11 01:10
I think the tree thing could be very novel. Especially if it looks like a tree...which I think is easier than the house. But we'll see!
nathaniel
2016-04-11 01:12
I agree
bruno-c
2016-04-11 01:12
should we like, answer him? or is he asking rhetorically?
jerclark
2016-04-11 01:13
In my mind it's very clear how the historical events will be linked to the views...but I guess the visualizations aren't clear enough. I think I might be the only one who sees it - which might be an indication of its effectiveness.
bruno-c
2016-04-11 01:13
i mean are we supposed to keep honing that project plan until the project is done :simple_smile:
jerclark
2016-04-11 01:13
I think its rhetorical
bruno-c
2016-04-11 01:13
well, doesn?t hurt to answer anyways hehe
nathaniel
2016-04-11 01:14
Also, it?s very hard to show interactions with static sketches
nathaniel
2016-04-11 01:14
I was impressed with how much Group 1 was able to understand though
bruno-c
2016-04-11 01:14
i think the historical event thing is just because of the lo-fi sketches. when we have the prototype it?ll be really obvious
bruno-c
2016-04-11 01:14
seriously yeah
jerclark
2016-04-11 01:15
OK...I can try to make that more obvious in the redeisgn.
bruno-c
2016-04-11 01:16
speaking of that of now we have enough data to prototype that idea
jerclark
2016-04-11 01:16
So, as far as a redesign - I think the expert evals + alain's comments point to the tree/house/whatever as needing the most attention.
jerclark
2016-04-11 01:16
The expert evals definitely had some good points that I think would be easy to integrate.
bruno-c
2016-04-11 01:17
> Use what you have learned about innovation in lecture to identify a weakness of your current design and propose an innovative solution (1 sketch)
jerclark
2016-04-11 01:18
There was a good question about individual vs. family spending
jerclark
2016-04-11 01:18
I'm not actually sure how the data is broken down?
bruno-c
2016-04-11 01:18
yeah. we can?t disaggregate
jerclark
2016-04-11 01:19
Its all household spending?
nathaniel
2016-04-11 01:19
Yeah, I think we will need to make that super clear
bruno-c
2016-04-11 01:19
yes
nathaniel
2016-04-11 01:19
Which means we can?t drill down on things like gender
jerclark
2016-04-11 01:19
OK/
bruno-c
2016-04-11 01:19
kinda weird that it?s all ?husband and wife?. what is this the 50s?
bruno-c
2016-04-11 01:20
?other husband and wife consumer units? no idea what that means hehe
bruno-c
2016-04-11 01:22
for me the tree visualization was also the big question mark, because i couldn?t conceive of it completely. if you want to stick with that one that?d be a good candidate for the ?innovative solution"
nathaniel
2016-04-11 01:23
I?m not sure I quite understand the ?small multiples? sketch at the end that group 1 provided
nathaniel
2016-04-11 01:23
Do either of you?
bruno-c
2016-04-11 01:23
i think he means show sparklines for many years
bruno-c
2016-04-11 01:23
inside the tree
jerclark
2016-04-11 01:24
I thought they meant sparklines for the 'subcategories'
nathaniel
2016-04-11 01:24
Oh ? I see
jerclark
2016-04-11 01:24
Right, that's what you're saying.
jerclark
2016-04-11 01:24
nvmd
bruno-c
2016-04-11 01:24
i guess it?s both!
bruno-c
2016-04-11 01:24
it?s sparklines all the way down
bruno-c
2016-04-11 01:25
i dunno, i think that?s too crazy
jerclark
2016-04-11 01:26
Perhaps.
jerclark
2016-04-11 01:26
I really believe in the tree.
jerclark
2016-04-11 01:26
The money tree.
jerclark
2016-04-11 01:27
The spending tree.
jerclark
2016-04-11 01:27
The giving tree.
nathaniel
2016-04-11 01:27
btw Bruno ? data set explorer is really useful. I feel like I finally understand what we have to work with
bruno-c
2016-04-11 01:27
joshua tree? tree of life
jerclark
2016-04-11 01:27
yes! yes!
jerclark
2016-04-11 01:27
But it should look like a tree.
bruno-c
2016-04-11 01:28
awesome. i was really bent on making it because otherwise it?s too hard to even see what?s available
jerclark
2016-04-11 01:28
bruno-c
2016-04-11 01:28
tbh it turned out to not be exactly what i thought but good enough
jerclark
2016-04-11 01:28
Agree - it's very helpful.
jerclark
2016-04-11 01:28
Thanks for putting it together.
bruno-c
2016-04-11 01:29
oh man if we could design a cool organic leaf drawing algorithm
bruno-c
2016-04-11 01:29
that?d be rad
jerclark
2016-04-11 01:29
We could do it...
jerclark
2016-04-11 01:29
Maybe not a leaf algo, but perhaps a 'sway' algo that shows how spending makes the tree bend.
nathaniel
2016-04-11 01:30
At some point the tree should fall over
jerclark
2016-04-11 01:30
And I don't think we need an algo...I think the right controls of the force layout in D3 could make this work.
jerclark
2016-04-11 01:30
There's gravity and resistance...might be all we need.
bruno-c
2016-04-11 01:30
i still have no idea what will be on the tree :simple_smile: sorry you explained it like 10 times, i?m just really slow
jerclark
2016-04-11 01:31
You're not slow.
jerclark
2016-04-11 01:31
Not even sure I totally have it...
jerclark
2016-04-11 01:31
But if you look at the cover of that book - it might be something like that, but a little cruder?
bruno-c
2016-04-11 01:31
you?ll be waking in the middle of night like ?eureka!"
bruno-c
2016-04-11 01:31
yeah i love the trees in that book
jerclark
2016-04-11 01:32
They look awesome
bruno-c
2016-04-11 01:32
the one on the cover on the right is nice
jerclark
2016-04-11 01:32
p 50
bruno-c
2016-04-11 01:32
really crafty looking
jerclark
2016-04-11 01:32
p52
jerclark
2016-04-11 01:33
but probably something more like p48
jerclark
2016-04-11 01:33
@jerclark uploaded a file: https://melodictransport.slack.com/files/jerclark/F0ZFZGVK4/screen_shot_2016-04-10_at_9.33.40_pm.png
bruno-c
2016-04-11 01:34
oh they talk about ernst haeckel too
bruno-c
2016-04-11 01:34
i have some his books
bruno-c
2016-04-11 01:34
i should go pull that out
nathaniel
2016-04-11 01:35
I think the hardest part for me with the tree is how do you communicate that certain leafs are *within* other leafs. These charts do a great job of showing connections, but what you really have are levels of detail.
bruno-c
2016-04-11 01:35
you could click a branch and then branches would grow out of it :simple_smile:
nathaniel
2016-04-11 01:36
That could work if you zoom in on the leaf
nathaniel
2016-04-11 01:36
or show levels with height
bruno-c
2016-04-11 01:37
ok, well let?s back up a little bit for my benefit
jerclark
2016-04-11 01:37
Here's my concept for the encoding:
bruno-c
2016-04-11 01:37
suppose we start from the data explorer. what do i pick from there to put in the tree? do i split up expenses on the left and right?
jerclark
2016-04-11 01:38
OK good question.
jerclark
2016-04-11 01:38
First, pick a year
jerclark
2016-04-11 01:38
Get the avg income for all respondents for that year.
jerclark
2016-04-11 01:39
That would be encoded in the size of the trunk, baselined against the average of all years.
jerclark
2016-04-11 01:40
So, the average income across ALL years would be a dotted line trunk, and the average income for the selected year would be the actual trunk.
nathaniel
2016-04-11 01:40
That part I get
jerclark
2016-04-11 01:41
(Everything adjusted to today's dollars for income, BTW)
nathaniel
2016-04-11 01:41
and then branches are expenses, correct?
jerclark
2016-04-11 01:41
OK. So now we have a trunk that is sized according to income for that year. Bigger tree, more income.
jerclark
2016-04-11 01:41
Yes. and the length of the branches represents % of income.
bruno-c
2016-04-11 01:42
avg for ALL respondents? okay
jerclark
2016-04-11 01:42
Yes.
jerclark
2016-04-11 01:42
Very simple to start. If we can filter to a specific set of respondents, that's fine too - but just to start.
nathaniel
2016-04-11 01:42
Right, but we?re going to have global filters (much like in the data set explorer)
jerclark
2016-04-11 01:42
Sure. That's fine - but start simple for now to get the concept
nathaniel
2016-04-11 01:42
So dodent the tree represent whatever slice of the dataset you?re looking at?
nathaniel
2016-04-11 01:43
Sure
jerclark
2016-04-11 01:43
I haven't really figured out whether we can slice the data like that - but if we can, then yes.
nathaniel
2016-04-11 01:43
(dodent = doesn?t)
jerclark
2016-04-11 01:43
So, average incomes of the slice being looked at - but for arguments sake, consider all to be the default state.
nathaniel
2016-04-11 01:44
Is are sizes of expenses represented by the size of the leafs?
nathaniel
2016-04-11 01:44
Or is that encoded in the length of the branch?
jerclark
2016-04-11 01:44
Yes...but one sec.
nathaniel
2016-04-11 01:44
(Sorry, I?ll be quiet)
jerclark
2016-04-11 01:45
No worries! The length of the first level branches represents the % of spending for that WHOLE category. If people spent 30% on housing and 10% on fuel, the housing branch would be approx 3 times longer than the fuel branch.
jerclark
2016-04-11 01:46
At the end of each category branch could be a 'leaflike image' of that thing (like a house or an oil can, whatever - make it fun).
jerclark
2016-04-11 01:46
That image would be the category 'node'
jerclark
2016-04-11 01:48
Then off the category node would be the subcategories, where each branch is equal length but the sizes of the subcategory nodes represent unit price of that expenditure as a %of income. Not the total spent but the unit price.
jerclark
2016-04-11 01:49
Then, off of each subcategory node would be a really small edge and node for each unit consumed (total spent/unit price)
bruno-c
2016-04-11 01:49
ok ? data point: there are 15 categories of spending
jerclark
2016-04-11 01:49
Right.
bruno-c
2016-04-11 01:49
that?s a good number
bruno-c
2016-04-11 01:49
it?s plenty of branches
jerclark
2016-04-11 01:49
Plenty, if not too much.
jerclark
2016-04-11 01:49
Could even have another interaction to choose specific categories...
jerclark
2016-04-11 01:50
But 15 might be just perfect.
bruno-c
2016-04-11 01:50
if you order the branches with the long ones at the bottom you could have kind of a conifer shape
bruno-c
2016-04-11 01:52
you know what if you boil it down ? a rough idea of what the tree would be like could be a bar chart that grows on both sides of a vertical axis
jerclark
2016-04-11 01:52
You could...I was thinking they might 'bend' toward the ground...where length of branch would have some gravitional effect as would size of the subcategory leaves and the clusters of consumption. Let me draw a quick sketch...
bruno-c
2016-04-11 01:54
yeah that part is really hard to imagine :simple_smile: it bends because the branch is so long
jerclark
2016-04-11 01:55
AND because the unit prices are high and there's a lot of units consumed...
bruno-c
2016-04-11 01:55
yeah about that :simple_smile: we dont? have those yet
bruno-c
2016-04-11 01:56
we have money spent per item
bruno-c
2016-04-11 01:56
best case scenario is that we can have the information for a FEW items
jerclark
2016-04-11 01:58
OK - so the tree could be sparse.
bruno-c
2016-04-11 01:59
unfortunately after looking at the data more attentively i feel like one of the questions we really wanted to answer regarding the spending behaviour
bruno-c
2016-04-11 01:59
well i?m not sure how we can answer that question
bruno-c
2016-04-11 02:00
re. are we spending more to keep up, or spending the same and consuming less
jerclark
2016-04-11 02:00
We would need to know the cost of things
jerclark
2016-04-11 02:01
We have enough to make it interesting, I think, no? Even if it's just housing, food and fuel.
jerclark
2016-04-11 02:01
For the tree, that is.
nathaniel
2016-04-11 02:02
There are also the data points in there about net-change in assets and liabilities
nathaniel
2016-04-11 02:02
it?s not exactly the same thing as showing the difference between income and spending
jerclark
2016-04-11 02:02
@jerclark uploaded a file: https://melodictransport.slack.com/files/jerclark/F0ZFWHPLZ/photo_on_4-10-16_at_10.02_pm.jpg
nathaniel
2016-04-11 02:02
but presumably assets go up when people are feeling flush
jerclark
2016-04-11 02:03
In this picture, the tree is weighed down by fuel costs (as evidenced by the 'hanging'), even though they may not be buying that much gas as evidenced by the cluster of units purchased.
nathaniel
2016-04-11 02:04
And just to be clear, you?re encoding area size in both area of the circle and length of the branch, correct?
jerclark
2016-04-11 02:04
So you can see if the tree hangs low, but there isn't a large cluster of stuff off the end...you're not getting a lot for your money.
jerclark
2016-04-11 02:04
That's definitely tricky.
jerclark
2016-04-11 02:04
Length of first level branch is total spending for that category...So food is less than housing is less than fuel (in this picture)
jerclark
2016-04-11 02:05
Size of leaf nodes is 'unit price of that thing'...so in this image unit price of most food items is small. Unit price of oil is less than gas.
jerclark
2016-04-11 02:06
Number of units purchased are the short scilia hanging off the leaf nodes.
jerclark
2016-04-11 02:07
This creates a 'clustering' effect to show how much of something people bought...if there's a bunch of stuff hanging off a small unit price, it will have a potentially similar 'weighting' effect to a few scilia hanging off of a large unit price.
jerclark
2016-04-11 02:07
As the clusters get smaller, you can see how much less people spent on something. Especially at a gist level. If fuel costs go up, the drag will still be low, the branch will be 'barren'
jerclark
2016-04-11 02:08
Or, if fuel costs go up and consumption stays steady or increases, the tree will appear very lopsided since the branch will be long, the leaves will be big AND the clusters will be big.
jerclark
2016-04-11 02:09
This may all be a bunch of hooey...but I just want to get across what I'm thinking. The visual concepts of bending and clustering can represent 'strain' on a budget' and 'appetite'
jerclark
2016-04-11 02:09
Maybe not appetite, but at least consumption.
nathaniel
2016-04-11 02:10
Yeah, I really like the bending idea
nathaniel
2016-04-11 02:10
And the income/width of a trunk idea
jerclark
2016-04-11 02:12
The bending can happen by using line interpolation built into D3 or by force layouts where gravity pulls things toward the ground and we set high repulsion between big categories and the trunk. I don't even think we'd need to draw lines...we can just set repulsion values in a force layout and the edges (branches) will draw themselves
bruno-c
2016-04-11 02:12
actually the income/width of the trunk i didn?t get the point of
bruno-c
2016-04-11 02:13
in terms of relative money, wouldn?t the average stay the same size by definition?
jerclark
2016-04-11 02:14
If we look at two trees side by side: 1990 and 2000, lets say, we may have been making more money (considered against the average of all recorded years) in 1990 than in 2000. I think it would be interesting to see the 1990 tree and ALL of the branches, leaves etc bigger than those in 2000.
jerclark
2016-04-11 02:14
It shows that there was 'more' money earned/spent in 1990 than 2000.
jerclark
2016-04-11 02:15
It's helpful - cuz lets say the tree in 1990 is much bigger, but all of the branches are stable (i.e. not bending) and consumption is very low. It would be interesting to see that just because we made more we didn't spend more proportionally.
jerclark
2016-04-11 02:16
And that would be obvious if the 1990 tree was bigger.
jerclark
2016-04-11 02:16
But I might be missing something in the concern you expressed.
bruno-c
2016-04-11 02:17
got it
jerclark
2016-04-11 02:17
It might be a complicated concept, but I think if we calibrate it right...the tree 1) could be very simple to implement and 2) be easy to interpret.
jerclark
2016-04-11 02:19
The one big concern I have (and that doesn't mean its the only valid concern!) is the fact that branch length encodes spending as % of income and leaf size encodes unit price as % of income. But basically, I wanted to encode the 'amount spent' in the first-level branches and the 'amount consumed' in the second level branches...but they will both have an impact on the 'drag'
bruno-c
2016-04-11 02:23
okay. i will try to find more data to allow that ?units? level to work
bruno-c
2016-04-11 02:24
but just to be on the safe side, maybe think of an alternate way to get a similar effect without the unit level
jerclark
2016-04-11 02:24
I think we could make a go with just the ap stuff we have now...food, housing and fuel are probably the most interesting cuz we all have to buy them.
jerclark
2016-04-11 02:25
I think it would be OK to have the trees only represent a subset of the stuff that's included in the area chart.
bruno-c
2016-04-11 02:25
yeah that makes sense.
bruno-c
2016-04-11 02:26
there?s data on here that?s pretty mind boggling, like sudden spikes in ?floor covering? spending across the board
jerclark
2016-04-11 02:26
That kind of thing is awesome. Maybe it comes after a flood?
bruno-c
2016-04-11 02:26
yeah it?s weird
jerclark
2016-04-11 02:26
Katrina? Sandy?
bruno-c
2016-04-11 02:26
and asian people eat less sweets, not terribly surprised
jerclark
2016-04-11 02:26
Nice.
bruno-c
2016-04-11 02:27
i think there was a big crash in 2003? there?s some crazy anomalies there too
jerclark
2016-04-11 02:27
Yes. and I'd imagine 2008 is interesting too.
jerclark
2016-04-11 02:28
In any case - thanks for entertaining my stream of consciounsness.
bruno-c
2016-04-11 02:28
it?s really interesting, it?s just that visually and conceptually it?s pretty packed
bruno-c
2016-04-11 02:28
kind of thing that totally makes sense to the person imagining it
bruno-c
2016-04-11 02:28
:simple_smile:
bruno-c
2016-04-11 02:29
something that happens with me and programming all the time
bruno-c
2016-04-11 02:30
anyways i think that?s it for tonight for me, i?ll keep on plugging some data in there tomorrow and then i also have to hustle on my other class homework. they keep giving us team problem sets
jerclark
2016-04-11 02:31
Cool :wink: I'm definitely game for other crazy ideas. Let me take a crack at a new tree sketch tomorrow, with more clear encodings, etc.
bruno-c
2016-04-11 02:31
cool. i do like the packed circle they put in the feedback too
bruno-c
2016-04-11 02:31
not nearly as original but it?s nice
jerclark
2016-04-11 02:32
So the redeisgn:
jerclark
2016-04-11 02:32
Use what you have learned about innovation in lecture to identify a weakness of your current design and propose an innovative solution (1 sketch)
Use the (peer) feedback you have received so far and the knowledge from the poster session to re-design 1-2 existing visualizations in your project
bruno-c
2016-04-11 02:32
a large ?concentric circles? visualization would show the percentage of income spent on categories really well too. ]
bruno-c
2016-04-11 02:32
like a multicolor jawbreaker
jerclark
2016-04-11 02:33
It would but I don't like color as much for quantitative data (although its arguably ordinal)...and there's not a good way to indicate amount spent, amount consumed and unit price. But it does look cool and could be a good alternative.
jerclark
2016-04-11 02:34
If I update the tree sketch (or do something different), what element of the redisgn assignment do you think it satisfies?
jerclark
2016-04-11 02:34
"lesson from innovation", or "peer feedback update"?
jerclark
2016-04-11 02:35
Maybe more the innovation side. Then any data updates or tweaks to what we had could be the 'peer feedback' updates.
jerclark
2016-04-11 02:35
Anyhow...I'm off to bed, maybe we chat tomorrow night again?
nathaniel
2016-04-11 02:35
Hmm?
nathaniel
2016-04-11 02:36
I?ll try and think about that Jeramy
bruno-c
2016-04-11 02:36
k
nathaniel
2016-04-11 02:36
but I would goo with peer feedback update
nathaniel
2016-04-11 02:36
because that?s what we?re responding too
jerclark
2016-04-11 02:36
OK...so we'd have to do something 'innovative' with another part of the viz, right?
bruno-c
2016-04-11 02:37
i think the tree is innovative enough?
nathaniel
2016-04-11 02:37
Yes, it is
bruno-c
2016-04-11 02:37
bruno-c
2016-04-11 02:37
@bruno-c pinned a message to this channel.
jerclark
2016-04-11 02:38
Damn, there is some crazy shit out there.
bruno-c
2016-04-11 02:38
this guy is the shit
jerclark
2016-04-11 02:38
Stop! :simple_smile:
bruno-c
2016-04-11 02:39
lol
jerclark
2016-04-11 02:39
For reals, though - I gotta head to bed, really early flight. Back in boston in AM, I will puzzle out some updates to the the tree....and you guys will pick off some other stuff?
jerclark
2016-04-11 02:40
BTW, if you think the tree thing is total shite, I'm totally open to other ideas.
jerclark
2016-04-11 02:40
For real.
nathaniel
2016-04-11 02:41
I think it can work. We may still have to make some modifications to it, and I?m not sure we?ll have all of the required data.
nathaniel
2016-04-11 02:41
But I like the bending/weighing down metaphor a lot
jerclark
2016-04-11 02:42
All right, I'm happy to keep prototyping if you think there's something in it...and harsh feedback is welcome.
bruno-c
2016-04-11 02:42
same here. it?ll all come down to how we execute it
jerclark
2016-04-11 02:42
Yes
jerclark
2016-04-11 02:43
Alright fellas, g'nite. Chat tomorrow.
nathaniel
2016-04-11 02:43
Good night
bruno-c
2016-04-11 02:43
later
nathaniel
2016-04-11 23:20
Hi guys ? just as an FYI, I?ve put our ?process book? document here:
nathaniel
2016-04-11 23:20
nathaniel
2016-04-11 23:21
Feel free to add things too it that you think would be helpful.
nathaniel
2016-04-11 23:25
Jeremy ? I?m happy to extract out the expert feedback ?entry? in a PDF if you want it for tonight's submission. Or you can just submit the whole thing.
jerclark
2016-04-12 00:35
I was thinking I?d submit the whole thing.
bruno-c
2016-04-12 00:42
i?m probably going to go through that at one point to add some stuff but i think that?s a lot of stuff for now
jerclark
2016-04-12 00:44
I was looking at the data, and I think a simple line chart that shows the delta between total money earned and total money spent. I think that data is in the data browser. That tool is awesome, BTW.
jerclark
2016-04-12 01:02
I?ve been looking at the tree all day. I?m about to become a lumber jack!
jerclark
2016-04-12 01:03
But for real, I think I?ve got something interesting?but I?m finding it hard to consider the ?unit? price of a housing.
jerclark
2016-04-12 01:07
Maybe housing isn?t really a ?consumer? thing?
bruno-c
2016-04-12 01:08
well, i guess we all need housing so it?s also a basic need
jerclark
2016-04-12 01:08
True.
jerclark
2016-04-12 01:08
But we can still ?consume? more housing than we need.
jerclark
2016-04-12 01:09
It?s ultimately a choice what type of housing we live in.
jerclark
2016-04-12 01:10
But while its easy to calculate ?average consumption? of eggs (Total spent/price per dozen == dozens consumed), it?s a little different to do housing because we don?t usually pay for it all at once.
jerclark
2016-04-12 01:10
If I spend 1000/month in mortgage, what am I really ?consuming??
bruno-c
2016-04-12 01:37
square feet?
bruno-c
2016-04-12 01:37
:simple_smile:
jerclark
2016-04-12 01:37
Right, of course!
bruno-c
2016-04-12 01:39
it gets tricky. asseen during the last housing crisis if you make housing cheaper and cheaper eventually people will consume beyond their means. it seems that culturally most people want more square footage than they have or can afford
bruno-c
2016-04-12 01:40
it?s a status symbol in many places, too
jerclark
2016-04-12 01:40
Right - I think it would be really cool to visualize that.
bruno-c
2016-04-12 01:40
Yes. Maybe in CS271 :wink:
jerclark
2016-04-12 01:40
what is 271?
bruno-c
2016-04-12 01:40
i?m just making that up heh
jerclark
2016-04-12 01:40
aha.
jerclark
2016-04-12 01:41
So what about this CPI stuff?
bruno-c
2016-04-12 01:41
oh yeah.
jerclark
2016-04-12 01:41
It?s not actually values, right? It?s a factor based on some baseline?
bruno-c
2016-04-12 01:41
it?s ?a basket of goods and services"
bruno-c
2016-04-12 01:42
i think you can get a breakdown on the website
bruno-c
2016-04-12 01:42
probably
jerclark
2016-04-12 01:42
But when I look at the values, what am I looking at?
bruno-c
2016-04-12 01:42
how much does the basket cost, month over month
jerclark
2016-04-12 01:42
For example, here?s shelter from 1997:
jerclark
2016-04-12 01:42
MUSR0000SA2 1997 M01 155.2
MUSR0000SA2 1997 M02 155.5
MUSR0000SA2 1997 M03 155.6
MUSR0000SA2 1997 M04 155.8
MUSR0000SA2 1997 M05 156.1
MUSR0000SA2 1997 M06 156.5
MUSR0000SA2 1997 M07 156.8
MUSR0000SA2 1997 M08 156.9
MUSR0000SA2 1997 M09 157.3
MUSR0000SA2 1997 M10 157.7
MUSR0000SA2 1997 M11 158.2
MUSR0000SA2 1997 M12 158.3
bruno-c
2016-04-12 01:42
so between jan and dec, it costs 3$ more to get that basket
bruno-c
2016-04-12 01:43
what, shelter?
jerclark
2016-04-12 01:43
Right- shelter.
bruno-c
2016-04-12 01:43
then i have no idea what that means
bruno-c
2016-04-12 01:43
hehe
jerclark
2016-04-12 01:43
I don?t think it really cost $158/month for shelter.
jerclark
2016-04-12 01:43
In 1997
bruno-c
2016-04-12 01:44
a tarp maybe
jerclark
2016-04-12 01:44
Mud hut. Yert
bruno-c
2016-04-12 01:44
with a sleeping bag
bruno-c
2016-04-12 01:44
heh
jerclark
2016-04-12 01:44
Yurt.
bruno-c
2016-04-12 01:44
i gotta try one of those vacations in a yurt
jerclark
2016-04-12 01:44
I think the idea is that there?s a baseline?in the case of shelter, it was 1982-1984, which is considered ?100? on the index.
bruno-c
2016-04-12 01:45
how about that
jerclark
2016-04-12 01:45
"Essentially, it
compares the cost of a sample "market basket" of goods and services in a
specific month relative to the cost of the same "market basket" in an
earlier reference period. This reference period is designated as the base
period."
bruno-c
2016-04-12 01:45
ooooh bananas
bruno-c
2016-04-12 01:46
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F0ZRJ3GJC/-.js
jerclark
2016-04-12 01:58
Thats the stuff that?s in the AP file I think.
jerclark
2016-04-12 02:03
From the BLS website: "It is unlikely, however, that the difficult problems of defining living standards and measuring changes in the cost of their attainment over time will ever be resolved completely."
jerclark
2016-04-12 02:04
So what are we going to submit for the redesign, besides perhaps my updated sketch?
jerclark
2016-04-12 02:04
Nothing has been done and it?s worth a larger amount of points for the ?project'
jerclark
2016-04-12 02:12
You guys around?
jerclark
2016-04-12 02:13
Wait do you mean submit the whole process book?
jerclark
2016-04-12 02:13
I don?t think we need to do that.
jerclark
2016-04-12 03:17
Nathaniel - did you want to add some feedback for Group 3?
jerclark
2016-04-12 03:29
I submitted some files for the redesign. I hope you?re OK with what I submitted?Maybe I missed a discussion, but I would have loved to talk a bit more about it.
nathaniel
2016-04-12 03:50
My understanding was that I was not supposed to submit feedback because I was not an ?evaluator?
nathaniel
2016-04-12 03:53
Sorry if I misunderstood!
nathaniel
2016-04-12 03:59
And thanks Jeremy for doing the submission. (I?ve been in transit for the last couple of hours.)
jerclark
2016-04-12 11:16
No worries. Have a look at what I submitted when you get a chance!
jerclark
2016-04-13 03:06
Hey guys?added some issues to github, just as a start: https://github.com/nathanielburbank/melodictransport/issues
jerclark
2016-04-13 03:06
Please add, edit, remove as you see fit!
jerclark
2016-04-15 01:06
Hey guys - just reaching out to see if you want to connect soon.
bruno-c
2016-04-15 01:08
yeah definitely
bruno-c
2016-04-15 01:08
we have a shit ton of things to do this weekend
bruno-c
2016-04-15 01:08
unless something has been happening while i wasn?t looking :simple_smile:
jerclark
2016-04-15 01:09
Yeah - there?s a bit to get done. I did put in a bunch of issues to track the work: https://github.com/nathanielburbank/melodictransport/issues
bruno-c
2016-04-15 01:10
ah cool
bruno-c
2016-04-15 01:11
so for me basically i am trying to wrap up some cs51 homework tonight to clear up tomorrow and weekend
bruno-c
2016-04-15 01:13
looking at the list of tasks, i don?t see any blockers for all visualization except the xmas money-tree which is the unique one. i?ve found many examples of radar for example
bruno-c
2016-04-15 01:14
i have got a selection of history events done but we could probably add more
jerclark
2016-04-15 01:14
That?s cool?how do you want to break down the work? I?m especially interested on your thoughts for #6 https://github.com/nathanielburbank/melodictransport/issues/6
jerclark
2016-04-15 01:14
I think the history events can be part of prototype 2?so long as we show that we can tie historical events into the charts.
bruno-c
2016-04-15 01:15
since it hasn?t been an issue so far to load everything upfront i?d suggest we wrap the data in some kind of simple api for querying
bruno-c
2016-04-15 01:15
i need to go back to previous homework to recall the data format that makes the most sense for d3 but basically all datasets are loaded
bruno-c
2016-04-15 01:17
i need to tie the mechanics for filtering that we have to the dataset in a way that?ll be easy to adapt for all the work i guess
jerclark
2016-04-15 01:17
I?m just thinking we can query the sqlite db and convert the results to JSON. Can we do joins in sqlite?
bruno-c
2016-04-15 01:18
yeah for sure
jerclark
2016-04-15 01:18
So, we just make queries from the sqlite DB you?ve built, convert result sets to JSON and load that into d3. Probably have 3 or 4 queries total.
bruno-c
2016-04-15 01:19
really? i guess i saw it as much more than that
jerclark
2016-04-15 01:19
Guess we need to figure that out!
bruno-c
2016-04-15 01:19
it depends on what we will offer to filter i guess
jerclark
2016-04-15 01:20
Filters can just be the same high-level queries with parameters
jerclark
2016-04-15 01:20
I think...
jerclark
2016-04-15 01:20
I?ll try to put together a sample for what I think could work.
bruno-c
2016-04-15 01:20
it?s a little tricky because of the way the data is structured but sure. i think i removed the sqllite database from the repo because we weren?t really using it and it?s huge
bruno-c
2016-04-15 01:20
but i can post it here
bruno-c
2016-04-15 01:22
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F10T11QP7/cx_copy.db.zip
bruno-c
2016-04-15 01:22
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F10ST3X3Q/cx_copy.db.zip
nathaniel
2016-04-15 01:22
Hi guys ? sorry was (still am) on a phone call
jerclark
2016-04-15 01:24
No worries - just an informal chat
nathaniel
2016-04-15 01:24
Okay. I?m also clearing my weekend to work on this.
nathaniel
2016-04-15 01:24
I?m going to suggest we keep our Saturday meeting
nathaniel
2016-04-15 01:25
Does 1 pm work for the two you still?
jerclark
2016-04-15 01:25
Should be good for me?.
nathaniel
2016-04-15 01:25
(Although I?m happy to chat tomorrow night if you?d prefer)
jerclark
2016-04-15 03:19
This query against your DB bruno will return a list of all values from the ?data? table with the series information expanded to show the category, subcategory, item name, demographic and characteristic information. It?s probably more than we need but I think would do most things we?ve discussed:
jerclark
2016-04-15 03:19
select
cat.category_text as 'Category',
sub.subcategory_text as 'Subcategory',
item.item_text as 'Item',
demo.demographics_text as 'Demographic',
characteristics.characteristics_text as 'Characteristics',
d.value as 'Value',
d.year as 'Year',
d.period as 'Period'
from 'cx.data' as d
join 'cx.series' as ser USING ('series_id')
join 'cx.demographics' as demo USING ('demographics_code')
join 'cx.characteristics' as characteristics ON ((ser.characteristics_code = characteristics.characteristics_code) AND (ser.demographics_code = characteristics.demographics_code))
join 'cx.category' as cat USING ('category_code')
join 'cx.subcategory' as sub USING ('subcategory_code')
join 'cx.item' as item USING ('item_code');
jerclark
2016-04-15 03:19
Unfortunately, it?s a big file: 55MB
jerclark
2016-04-15 03:21
We could probably just load the various files like you mentioned and create a little API around the data. I?ll have to play around with it a little more.
bruno-c
2016-04-16 00:10
yeah, that?ll get you every data item.
bruno-c
2016-04-16 00:11
i think what we should do is probably take each visualization and just detail what data is needed, grouped in what way
jerclark
2016-04-16 00:11
But it?s a lot of redundant text in repeated demographic and characteristic names. Probably better to related them with the ids and create a little API around them.
bruno-c
2016-04-16 00:11
the way that the series are designed in cx we?ll have to aggregate things sometimes
jerclark
2016-04-16 00:11
How do you want to divvy up the work?
bruno-c
2016-04-16 00:12
if you tell me here?s how i want the data to look for my visualization, i?ll make that happen and then i can take on some of the viz myself
bruno-c
2016-04-16 00:13
we should probably flesh out the ?interaction diagram? i guess with the corresponding data
bruno-c
2016-04-16 00:13
especially if we won?t make the entire thing available
jerclark
2016-04-16 00:14
OK, that?s cool. I?ll work on that after I get the kids to bed.
bruno-c
2016-04-16 00:15
awesome. i still have a bit more cs51 to wrap up but after that i?ll work on the prototype
jerclark
2016-04-16 00:15
No worries.
jerclark
2016-04-16 02:12
For the stack chart, I think something like this will work (basically a lift from the stacked area lab assignment):
jerclark
2016-04-16 02:13
{
"incomes": [
{
"year": "1984",
"income": 20000
},
{
"year": "1985",
"income": 21000
}
],
"categories": [
{
"category": "Housing",
"values": [
{
"year": "1984",
"value": 10000
},
{
"year": "1984",
"value": 11000
}
]
}
],
"expenditures": [
{
"expenditure": "Flooring",
"category": "Housing",
"values": [
{
"year": "1984",
"value": 100
},
{
"year": "1985",
"value": 110
}
]
}
]
}
jerclark
2016-04-16 02:13
You can see the whole schema here:
jerclark
2016-04-16 02:15
I think this schema works - if we want to ?filter? the data based on demographics or other characteristics, I think that?s when a little api against the SQLite db will come in handy.
jerclark
2016-04-16 03:12
For Radar chart, possible schema and sample here: http://tinyurl.com/jdu2tgg
jerclark
2016-04-16 03:13
We may be able to expand this to have an array of demographics at the top level and include all demographic breakdowns in 1 file.
nathaniel
2016-04-16 17:02
Hi Guy ? happy to chat now, or later this afternoon. Let me know what works for the two of you.
jerclark
2016-04-16 17:08
I can join in a couple minutes...it's a good time for me now.
nathaniel
2016-04-16 17:10
(Bruno and I are chatting now, but no rush. Join when you?re ready?)
jerclark
2016-04-16 19:07
Do a quick pull from github - I added the js and css libs
nathaniel
2016-04-16 19:07
Great ? will do
bruno-c
2016-04-16 19:15
cool i added the rest of the libs
bruno-c
2016-04-16 22:08
okay i got some pretty good data thing going now
bruno-c
2016-04-16 22:08
here are some things you can do ? i put these examples in the main.js
bruno-c
2016-04-16 22:10
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F119H12KA/-.js
bruno-c
2016-04-16 22:10
hmm, i think i?ll clean the values so that it?s only the value instead of an object, that would be simpler
nathaniel
2016-04-16 22:11
Awesome!
bruno-c
2016-04-16 22:11
the coolest thing though is
bruno-c
2016-04-16 22:11
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F119K7AMT/-.js
bruno-c
2016-04-16 22:13
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F119BARDL/-.js
bruno-c
2016-04-16 22:13
so if we have say a form where you can pick many items, you could group them together into a single ?slice? on the graph, for example
nathaniel
2016-04-16 22:14
So you just end up with a list rather than need to access each year individually?
bruno-c
2016-04-16 22:15
ah yes you?re right, in this case i forgot to include the year, missed that. i?ll make the values array be { year: year, value: value }
bruno-c
2016-04-16 22:15
in all cases
nathaniel
2016-04-16 22:16
yeah, that makes sense
bruno-c
2016-04-16 22:43
okay, i think the format is good now
bruno-c
2016-04-16 22:44
the values array always contains values in the form of
bruno-c
2016-04-16 22:44
`{ year: 1999, value: 23 }`
bruno-c
2016-04-16 22:44
the merged datasets, if we end up using it, are the same except they don?t include the text. the text is there to make updating labels easier basically
bruno-c
2016-04-16 22:46
as a reminder (it?s not a terribly popular JS feature) if you have have an array of things that you want to pass to a function as if they were multiple arguments
bruno-c
2016-04-16 22:46
you?d do something like
bruno-c
2016-04-16 22:46
`ds.for.apply(ds, [criteria1, criteria2]);`
bruno-c
2016-04-16 22:47
like if your criteria are dynamic basically
bruno-c
2016-04-16 22:47
but if you know what you have in advance, it?s easier to do `ds.for(criteria1, criteria2)`
bruno-c
2016-04-16 22:48
maybe i should rename `for` for query, since for is a reserved word. maybe later. bunch of costco shit to unpack and put away, i?ll be back in a little while
jerclark
2016-04-17 00:17
This looks really awesome?I?ll definitely bang around on it! I?ll spend some time tonight, but my wife and kids are heading out of town tomorrow afternoon, so I?ll have some uninterrupted time tomorrow to keep cranking.
bruno-c
2016-04-17 02:12
update: working on incorporating relative values ? dollars adjusted for inflation and % of pretax / after tax income
bruno-c
2016-04-17 03:07
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F11A16NLF/-.js
bruno-c
2016-04-17 03:08
so now the results included dollars adjusted for 2014 as well as the value in terms of percentage of income for the given demographic (based on before taxes)
bruno-c
2016-04-17 03:08
probably should do after taxes, i dunno
bruno-c
2016-04-17 03:09
merged query results also include the relative values
bruno-c
2016-04-17 03:11
think that?s it for me today, we can catch up tomorrow
bruno-c
2016-04-17 15:48
i see some added and changed files! no updates here though!
nathaniel
2016-04-17 15:53
Yeah ? still working on my part.
nathaniel
2016-04-17 15:54
Almost got a working version
bruno-c
2016-04-17 15:54
yeah is see it?s pretty close!
bruno-c
2016-04-17 15:55
the itemdatafordemographics, can you quickly explain what it?s for? is is to have all the the different characteristics as ?groups? for a given demographic?
jerclark
2016-04-17 15:55
Radar chart is working with fake data will commit in a little bit to use real data.
bruno-c
2016-04-17 15:57
i?m thinking of creating the ?event scrubber? maybe next
nathaniel
2016-04-17 16:00
Data question ? is there currently a way to build an array of sub categories for a given without querying for each of them individually?
bruno-c
2016-04-17 16:01
subcategories are under the ?category_code? ? is that what you mean?
bruno-c
2016-04-17 16:01
or items?
bruno-c
2016-04-17 16:01
category -> subcategory -> items
nathaniel
2016-04-17 16:02
Right, I mean items
bruno-c
2016-04-17 16:02
sure, you mean just the item codes and names?
nathaniel
2016-04-17 16:03
Currently, to build my data object, I?m doing something like this: var basic_expends = ds.query({
name: "Housing",
item: "HOUSING"
},{
name: "Healthcare",
item: "HEALTH"
}, etc?)
nathaniel
2016-04-17 16:04
I can do that for the items, as well, or I make list and then iterate through it
nathaniel
2016-04-17 16:04
It?s just tedious.
bruno-c
2016-04-17 16:06
i?m not sure i understand- is that an arbitrary list of items or are they all members of the same subcategory?
nathaniel
2016-04-17 16:11
Currently, I?m just working with sub-categories, but I will soon need to go get all of the items for each sub-category. Basically I?m just asking if there?s currently a method to return an array of the all of the item names for a given sub-category. I can write one, just trying to not duplicate work.
bruno-c
2016-04-17 16:11
i have one now, i?ll commit it
bruno-c
2016-04-17 16:16
okay, `ds.items()` to get every item code, `ds.items(?HEALTH?)` to get all item codes for category health. return format is `{ name: ?the name?, item: ?the code?, subcategory: ?the subcategory code? }`
nathaniel
2016-04-17 16:19
great ? thanks!
jerclark
2016-04-17 16:26
ItemsForDemographic is as you thought...for a particular item, get it grouped by every characteristic for a specific demographic. Its a similar concern to NB's. I envision users picking an item (probably from the percent distributions) and a demographic, then having a radar chart display a spoke for each characteristic. The radar chart scales will update for the given range and number of dimensions, so it should be able to handle user input quite easily.
bruno-c
2016-04-17 16:27
sounds good
jerclark
2016-04-17 16:29
Can be renamed/refactored as we see fit. Maybe items() can take three parameters: an item name, a group type, and a group id - then it's reusable for any type of grouping.
jerclark
2016-04-17 16:29
But may not be worth the trouble.
bruno-c
2016-04-17 16:29
group type?
bruno-c
2016-04-17 16:29
it already takes subcategory though
jerclark
2016-04-17 16:33
Yeah, maybe doesn't make sense. Subcats and items are in same hierarchy.
bruno-c
2016-04-17 16:34
once you got the radar in there i?ll probably change characteristics to work the same way as items, for example
jerclark
2016-04-17 16:34
Forget that suggestion.
bruno-c
2016-04-17 16:34
i tried to premptively guess at what you guys may need
bruno-c
2016-04-17 16:34
looks like it wasn?t too far off which is good!
jerclark
2016-04-17 16:36
OK. I'll start working in about an hour. You can look at radar in the meantime to see how I've thought about it's
bruno-c
2016-04-17 16:36
okay. so no uncommittted changes?
bruno-c
2016-04-17 16:36
just want to make sure i don?t clobber something
jerclark
2016-04-17 17:25
no uncommitted changes...
jerclark
2016-04-17 17:25
I?m working on things now.
bruno-c
2016-04-17 17:26
K give me a sec I started something
jerclark
2016-04-17 17:26
OK - take your time.
bruno-c
2016-04-17 17:26
Well it touches the radar
jerclark
2016-04-17 17:26
Sure?I?ll hold off.
nathaniel
2016-04-17 17:26
I have uncommitted changes, but only in stacked-area.js
bruno-c
2016-04-17 17:26
K
nathaniel
2016-04-17 17:26
(So I can merge easily)
jerclark
2016-04-17 17:27
Haven?t used _.js chain method?will look at that.
jerclark
2016-04-17 17:33
stepping away for a few
bruno-c
2016-04-17 17:50
alright there?s a functioning BEEF RADAR
bruno-c
2016-04-17 17:50
in the index page
bruno-c
2016-04-17 17:51
gotta step out also, be back ~ 30m
bruno-c
2016-04-17 19:52
@nathaniel any changes you want to push to area chart? i want to start integrating some stuff together but i have to modify the chart so that?s not bound to specific elements
nathaniel
2016-04-17 19:53
Yes, but I?m stepping away for a few minutes
nathaniel
2016-04-17 19:53
So, have at it
bruno-c
2016-04-17 19:53
Ok - if you can push the changes it?ll be easier when you have to merge back but it?s up to you
nathaniel
2016-04-17 19:54
(I?ve pushed my latest changes)
bruno-c
2016-04-17 19:54
ok got it
bruno-c
2016-04-17 20:01
stacked area chart in percent should be called a cake chart
bruno-c
2016-04-17 20:04
ah you smushed my changes it seems
bruno-c
2016-04-17 20:09
yay for git. recovered
bruno-c
2016-04-17 20:40
update: i moved the css for the charts into their own files, imported by the styles.css file. the index.html now shows both a radar and an area chart
bruno-c
2016-04-17 21:20
other thing i did is replace Dataset.prototype.itemDataForDemographic for queryDemographic, taking a criteria object as an argument just like query to keep things consistent
bruno-c
2016-04-17 21:20
instead of taking a prefix the names follow the format LB04-01-Characteristic Text
jerclark
2016-04-17 21:27
I saw that...
jerclark
2016-04-17 21:27
Looks good. Thanks.
bruno-c
2016-04-17 21:28
let me know when you want me to integrate your branch
jerclark
2016-04-17 21:28
You can integrate now - it still needs some work, but it?s functional and you can look at it.
jerclark
2016-04-17 21:28
It now has: data points on the spokes and hover tooltips
bruno-c
2016-04-17 21:29
cool
jerclark
2016-04-17 21:29
And First category is at noon now. Ring labels will be halfway between noon and the next spoke.
jerclark
2016-04-17 21:29
Probably need to shorten those labels!
bruno-c
2016-04-17 21:29
yeah i?ve been thinking of how we can deal with that, have a couple of ideas
jerclark
2016-04-17 21:29
Tooltip text is basic for now.
jerclark
2016-04-17 21:30
Should we add a ?display name? value to the raw data?
bruno-c
2016-04-17 21:30
yeah we could do that - we don?t actually need the raw name for anything so we can also just edit that
bruno-c
2016-04-17 21:31
just need to keep the codes intact
jerclark
2016-04-17 21:31
OK. If we?re not keying off it, that?s probably the best bet.
jerclark
2016-04-17 21:31
I?m now going to dropdowns to switch items and demographics.
bruno-c
2016-04-17 21:31
@bruno-c set the channel topic: beef radar
bruno-c
2016-04-17 21:31
okay
jerclark
2016-04-17 21:31
:simple_smile:
bruno-c
2016-04-17 21:32
you can get the list of demos, etc from ds now
jerclark
2016-04-17 21:32
As much as people will want to stick to the beef radar, we should let ?em switch.
bruno-c
2016-04-17 21:32
let?s just change the whole story to the story of beef consumption
jerclark
2016-04-17 21:32
Maybe you can think about some style updates? I think it could benefit from your eye.
jerclark
2016-04-17 21:33
It?s incredible how much beef middle aged people eat!
jerclark
2016-04-17 21:33
The style is lifted mostly from a polar plot gist on github.
bruno-c
2016-04-17 21:33
yeah, although really there?s a million things left to do
jerclark
2016-04-17 21:33
Sure?it?s not urgent.
bruno-c
2016-04-17 21:34
if you can, make your dropdowns into objects themselves
bruno-c
2016-04-17 21:34
that way we can inject them where we need on the page or have many of them
jerclark
2016-04-17 21:35
Sure.
bruno-c
2016-04-17 21:35
same way we do the vis i mean.. with a constructor, etc. one thing i did before that was really useful is using _.template
bruno-c
2016-04-17 21:35
for stuff like that its sooo fast
jerclark
2016-04-17 21:35
_.template?
bruno-c
2016-04-17 21:36
yeah you can do like
jerclark
2016-04-17 21:37
I?ll probably make a ?demographic? picker and an ?item? picker then.
bruno-c
2016-04-17 21:37
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F11BXEXL4/-.html
jerclark
2016-04-17 21:38
Is this an underscore.js thing?
bruno-c
2016-04-17 21:38
then you can do `_.template ( $(?#somedropdown?).text(), { choice: [ { name: ?choice1?, value : ?vlah? } ] })`
bruno-c
2016-04-17 21:38
yeah
jerclark
2016-04-17 21:38
Let me take a look at the docs.
bruno-c
2016-04-17 21:39
yeah it?s really really useful, i?m probably going to use for other parts of the screen
bruno-c
2016-04-17 21:39
much better than building html in strings for example
bruno-c
2016-04-17 21:40
i?m thinking of doing the event timeline ?controller"
bruno-c
2016-04-17 21:41
but i?ll need to consult you on that for sure
jerclark
2016-04-17 21:42
sure?let?s touch base in a bit.
bruno-c
2016-04-17 22:06
radar branch integrated
bruno-c
2016-04-17 22:07
i?m going to dinner, will be back in ~ 1.5h
bruno-c
2016-04-18 00:38
holy cow something is taking a really long time to load now
nathaniel
2016-04-18 00:39
It?s possible it?s the function I just added to look for layers with missing years and fill them in? it?s not exactly efficient
bruno-c
2016-04-18 00:39
ah yes
nathaniel
2016-04-18 00:40
the d3 stacked layout needs every layer to be the exact same length
bruno-c
2016-04-18 00:41
right --
bruno-c
2016-04-18 00:42
so in those cases you?re putting zeroes in there
nathaniel
2016-04-18 00:43
Right, which works for that layout. It might be simpler to have all queries return value arrays of the same length as an invariant on the data side. But I?m not sure if that works for the other visualizations, I and I just trying to get something working.
bruno-c
2016-04-18 00:44
got it
bruno-c
2016-04-18 00:44
yes we can fix that ahead of time in the dataset later
bruno-c
2016-04-18 00:45
we can also thin out the dataset to only keep 84-2014 data and remove what doesn?t fit
nathaniel
2016-04-18 00:46
It may also be slow because displaying every single item as it?s own layer may just be two many; but I wanted to try it.
nathaniel
2016-04-18 00:47
I need to step away for a few minutes. But perhaps we should all check in around 9:30?
jerclark
2016-04-18 00:47
Sounds good to me.
nathaniel
2016-04-18 00:47
k
bruno-c
2016-04-18 00:48
ya
bruno-c
2016-04-18 00:48
@nathaniel: are you playing in main.js right now?
nathaniel
2016-04-18 01:14
I was, but I have not done anything there that I actually want to save since my last push
nathaniel
2016-04-18 01:34
Hi guys ? I?m ready whenever you you would like to chat
jerclark
2016-04-18 02:51
defaults for radar chart:
jerclark
2016-04-18 02:51
this.options = _.defaults(options, {
width: 800,
height:800,
margin:{top: 40, right: 40, bottom: 40, left: 40},
showLabels:true
});
jerclark
2016-04-18 02:51
new Radar(parentElement, options)
jerclark
2016-04-18 02:51
notice, there?s no data since we?ll get that from the selector filters.
bruno-c
2016-04-18 02:53
are you sure? it?s not the radar itself that will listen for filter changes, it should be the radar ?controller?. but it?s not a big deal we can figure it out later
bruno-c
2016-04-18 03:01
okay so what i did is very simple
bruno-c
2016-04-18 03:02
the code that you have in wrangledata that loops through the keys (`Object.keys(_data)?`) i moved to dataset.js as ?toDimensions"
bruno-c
2016-04-18 03:02
queryDemographic will now return the values for all the years in a key called ?values"
bruno-c
2016-04-18 03:03
so when you `.data()` your accessor can go in the `values` and fetch the value you want (`_.where({ year: selectedYear })` is nice and quick)
jerclark
2016-04-18 21:55
Thanks BC...
bruno-c
2016-04-19 00:24
hey guys. i see some new branches! how are you guys doing
nathaniel
2016-04-19 00:25
I?m doing okay ? I probably need another hour or so to get the area chart into some sort of reasonable state.
nathaniel
2016-04-19 00:26
Trying to get the color pallets to make sense
jerclark
2016-04-19 00:26
Same, here?working on multiple path plots in the radar chart.
jerclark
2016-04-19 00:26
Not planning to make a lot of head way on the tree.
jerclark
2016-04-19 00:26
Does that seem OK to you guys?
nathaniel
2016-04-19 00:27
Yes, I think so. We will need to merge everything and turn it all back on so we have a version to submit.
jerclark
2016-04-19 00:27
Sure.
jerclark
2016-04-19 00:27
They gave us 10/10 on the redesign, so I?m not ?too? worried about grading?it?s more about how we feel about our progress!
nathaniel
2016-04-19 00:29
We don?t need to fix this tonight, but one problem I noticed this evening is that the ?percent of income? numbers do no add up. Or, they add up to more than 100% when you sum all of the expense items.
nathaniel
2016-04-19 00:29
@nathaniel uploaded a file: https://melodictransport.slack.com/files/nathaniel/F11N7JBSA/pasted_image_at_2016_04_18_08_29_pm.png
jerclark
2016-04-19 00:29
Sounds good. Put an issue in github?
nathaniel
2016-04-19 00:30
Yes, will do.
nathaniel
2016-04-19 00:30
It may just be that I?m not understanding what that item actually represents...
bruno-c
2016-04-19 00:33
yeah what i?m seeing on piazza is essentially submit what you have, give your TA something that shows progress
bruno-c
2016-04-19 00:37
i can do the integration whenever you guys feel ready then
bruno-c
2016-04-19 00:37
i?m gonna try to scare up some content / story
bruno-c
2016-04-19 00:44
some inspiration here for housing prices: http://www.economist.com/blogs/graphicdetail/2015/11/daily-chart-0
jerclark
2016-04-19 00:45
I was thinking of small multiple plots in rows like the US flag, with the ?legend? plot in where the blue section is.
bruno-c
2016-04-19 00:49
it?s too bad the regional demographics are such large groups. it?s really hard to show the insane climb in housing in urban centers
jerclark
2016-04-19 00:50
There?s urban v. rural somewhere I think?
bruno-c
2016-04-19 00:50
ooh yeah trying that
bruno-c
2016-04-19 01:03
yeah nope. doesn?t seem to want to give me the numbers i want
bruno-c
2016-04-19 01:03
it could just be that boston, sf, etc are outliers
bruno-c
2016-04-19 01:04
so add a bunch of cities with normal house inflation and it balances out. i don?t know enough about the mortgage situation in the us to know
bruno-c
2016-04-19 01:07
alright i have to start merging your branches in
bruno-c
2016-04-19 01:10
@jerclark: i think you forgot to check in your itempicker
nathaniel
2016-04-19 02:09
Okay ? I?m about ready to stop for the night?
nathaniel
2016-04-19 02:13
Bruno ? I can let you do the actual merge if you want, or should I just push to master? (I don?t think I have any conflicts)
bruno-c
2016-04-19 02:22
i?m doing the merges as i go already
bruno-c
2016-04-19 02:22
push to your branch. i?m missing the timeline in particular
bruno-c
2016-04-19 02:23
are we more clear on what we will give them? a link to the source code on github?
nathaniel
2016-04-19 02:36
Everything seems to indicate that they want a zipped project folder.
nathaniel
2016-04-19 02:36
(Even though that seems strange to me.)
bruno-c
2016-04-19 02:36
¯\_(?)_/¯
nathaniel
2016-04-19 02:36
(Timeline is part of my branch)
bruno-c
2016-04-19 02:37
yeah i don?t think it?s been pushed
bruno-c
2016-04-19 02:37
ah nvm
bruno-c
2016-04-19 02:37
i see it
nathaniel
2016-04-19 02:39
I did not spend much time on that today; just wanted to make sure the brushing still worked. I can takin on adding in some of our events data sets next?
bruno-c
2016-04-19 02:42
cool. yeah things still work after integration
bruno-c
2016-04-19 02:42
we should be okay for now with the stacked chart
nathaniel
2016-04-19 02:43
Yeah?I still need to add a legend and make the transitions make more sense. Currently, when you double click to zoom into a sub-catagory it?s kind of strange.
nathaniel
2016-04-19 02:44
But it works.
bruno-c
2016-04-19 02:49
yeah! i tried it and was like yasssssss
bruno-c
2016-04-19 02:49
i?m zooming into the cake
nathaniel
2016-04-19 02:56
Okay ? I?m going to go to bed soon. Jeramy ? are you good to do the actual submission again?
jerclark
2016-04-19 03:02
Sure.
jerclark
2016-04-19 03:02
I?m about to checkin a radar that updates with the timeline brush.
nathaniel
2016-04-19 03:06
Okay. Also, fyi, I added another entry to our process book as well...
jerclark
2016-04-19 03:21
Nice. I?m going to merge our changes onto master.
bruno-c
2016-04-19 03:25
i got a bunch of changes on tap here
bruno-c
2016-04-19 03:25
i merged all of nath?s stuff already
bruno-c
2016-04-19 03:25
and yours
jerclark
2016-04-19 03:28
OK - master should be all good to go
bruno-c
2016-04-19 03:29
okay, i will attempt to merge your new master with mine and then push up
jerclark
2016-04-19 03:30
The new radar that gets init?d in main will display two plots - one each for the beginning and end of the brushed area
jerclark
2016-04-19 03:30
Even ?animates? a little as you brush?
bruno-c
2016-04-19 03:31
using the timeline that nathaniel put in there before?
jerclark
2016-04-19 03:31
Yep
jerclark
2016-04-19 03:31
It?s all hooked up.
bruno-c
2016-04-19 03:32
k this merge is bananas but i?ll manage
jerclark
2016-04-19 03:32
Let?s chat for a sec...
jerclark
2016-04-19 03:32
share your screen with me.
bruno-c
2016-04-19 03:32
after that i?ll push and maybe have a quick look
bruno-c
2016-04-19 03:32
k
2016-04-19 03:33
@jerclark has started a Google+ Hangout for this channel. https://plus.google.com/hangouts/_/7srizmm7ifaw3o42i573quubhee.
jerclark
2016-04-19 03:37
@jerclark uploaded a file: https://melodictransport.slack.com/files/jerclark/F11P9ACFP/-.js
bruno-c
2016-04-19 03:45
nathaniel
2016-04-19 14:26
Just looking over the radar chart stuff now?really cool!
bruno-c
2016-04-21 01:18
doing some housekeeping in the dataset to make everything faster
bruno-c
2016-04-21 01:19
if you?ve changed things in the way your visualization expects data let me know
jerclark
2016-04-21 02:41
Not since monday night?s submission?.
bruno-c
2016-04-21 03:05
10sec -> 221ms now :simple_smile:
jerclark
2016-04-21 03:05
NICE!
jerclark
2016-04-21 03:09
BC, I was hoping to get CU (consumer price index) data for the tree?Do you think the current dataset model could be replicated for that? Or could we add CU data to the current implementation?
jerclark
2016-04-21 03:09
I know that?s asking a lot, just trying to see if you have any thoughts.
bruno-c
2016-04-21 03:14
i can definitely do that, i just haven?t studied that dataset yet. but yeah for sure
bruno-c
2016-04-21 03:15
i mean truth be told if i remain on data and layout / design duties i can do more with the data. i just figured that i would be needed to build more vis stuff
bruno-c
2016-04-21 04:03
alright so i did the data speed optimization for now but it looks like i broke the area chart a bit, the colors are off now and the label is a bit wonky. i will fix that tomorrow
nathaniel
2016-04-22 22:03
Hi Bruno and Jeremy ? I?m going to suggest again that we chat around 1 pm tomorrow to map out what we?d like to accomplish over the weekend for V2. Does that work for the two of you?
jerclark
2016-04-22 22:03
Yep
nathaniel
2016-04-22 22:03
great
bruno-c
2016-04-23 12:57
:+1:
2016-04-23 17:02
@bruno-c has started a Google+ Hangout for this channel. https://plus.google.com/hangouts/_/pbi32343vvdjjaaaisqoebil5me.
nathaniel
2016-04-23 17:02
Joining now...
nathaniel
2016-04-24 17:50
Hi guys ? I?m stopping for the afternoon.
nathaniel
2016-04-24 17:50
Several notes for Bruno. First, but the objects currently returned by ds.querySingle(c) include null subcategory properties; you?ll see I overwrite it in main.js (in my branch) but if that get?s fixed we can remove that line.
nathaniel
2016-04-24 17:51
Second, is there a way to query for items within expends only? (i.e., not including the income items and other asset items) Currently, I just hard-coded a list of subcategory titles to filter for, which works, but I?m guessing there?s a better way.
nathaniel
2016-04-24 17:52
Third, it would be helpful if the subcategory_text field could be included in the return object so we can easily map (for example) ?ALCBEVG? to ?Alcoholic beverages?, where we have room for the full title.
jerclark
2016-04-24 22:14
Crud - 15/20 on the prototype v1
jerclark
2016-04-24 22:42
Hey Bruno - have you taken a crack at any of the data stuff we talked about yesterday?
bruno-c
2016-04-25 00:19
Hey guys. I had a lot to do this weekend, I?ll be working this evening until quite late
jerclark
2016-04-25 00:21
No worries?busy times!
jerclark
2016-04-25 00:21
I?ll be online tonight too!
jerclark
2016-04-25 00:21
Been researching ideas for physical simulations with the tree?I?ve shelved that for static viz for now.
bruno-c
2016-04-25 00:24
Sounds reasonable.
jerclark
2016-04-25 00:26
I also think that the self-named item for a subcategory (i.e. FOODTOTL > FOODTOTL) is in-fact the survey result for total spent on food. If you look at the ?display_level? attribute, it?s 0 for all of those items?That means it should be at the top-level in an outline display. Why would they have a value for those items if it wasn?t valid?
bruno-c
2016-04-25 00:28
Yes, makes sense.
bruno-c
2016-04-25 00:29
It just doesn?t seem to add up in all cases but maybe there?s something wrong in my calculations
jerclark
2016-04-25 00:30
Maybe?I?m looking for some details on those number.
bruno-c
2016-04-25 01:11
Alright I merged nathaniel's changes, building a todo list based on our discussion yesterday and the feedback from today
jerclark
2016-04-25 01:18
K
bruno-c
2016-04-25 03:37
alright nathaniel - 1. i fixed the subcategory thing. 2. yes, we can query subcategories now, by category (e.g EXPEND). I updated main. 3. done
bruno-c
2016-04-25 03:38
i still have to do the more advanced grouping queries, just running out of time tonight
bruno-c
2016-04-25 03:39
i created a "floating" header with the timeline and value format selector. i'm thinking that we could basically have this floating filtering "palette" for the whole page such that all filtering is reflected (in different ways) in the vis
jerclark
2016-04-25 03:41
That would be awesome...
jerclark
2016-04-25 03:41
I?m almost done with a tree viz?I?ll share shortly.
bruno-c
2016-04-25 03:41
also for the layer cake chart, my guess is that when it's zoomed out, you don't actually want slices for every item - rather one slice per subcategory. after i do the data aggregating query thing i'll update it to do that
jerclark
2016-04-25 03:41
Will need to clean up tomorrow.
bruno-c
2016-04-25 03:41
cool. yeah def. have a lot of work ahead still for tomorrow night. ungh. just bad timing this week.
nathaniel
2016-04-25 03:42
Thanks Bruno!
nathaniel
2016-04-25 03:43
Will review in the am tomorrow morning?
bruno-c
2016-04-25 03:44
cool
jerclark
2016-04-25 04:11
@jerclark uploaded a file: https://melodictransport.slack.com/files/jerclark/F13B2BN9K/pasted_image_at_2016_04_25_12_11_am.png
jerclark
2016-04-25 04:12
Still working on some things - obviously - but figured I?d share.
bruno-c
2016-04-25 11:29
Awesome!
bruno-c
2016-04-25 11:29
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F13DA30DN/food_for_thought.jpg
jerclark
2016-04-25 14:05
A cool way to experience the 'tree' is to brush and/or drag along on the time line. If you brush to the right, you can really see how things change over time.
nathaniel
2016-04-25 18:44
Tree looks great! Really good work.
nathaniel
2016-04-25 18:44
We might want to move the timeline down lower on the page (or add an additional, connected scrubber) so you can move it and look at the tree at the same time? Not sure what would look best...
jerclark
2016-04-25 18:45
Bruno did that floating scrubber thing...do you like that or no?
nathaniel
2016-04-25 18:46
Oh, I see
nathaniel
2016-04-25 18:46
I had not notice that
nathaniel
2016-04-25 18:46
not noticed
nathaniel
2016-04-25 18:47
I guess that works
nathaniel
2016-04-25 18:48
I think we need to add year labels to all three charts, so it?s clear what you?re viewing
nathaniel
2016-04-25 18:49
(I will try adding that to the scrubber/timeline)
jerclark
2016-04-25 18:52
Definitely...I have a punch list of things to fix. I think the biggest thing to do at this point is craft a 'story' around the presentations. I'm thinking something like a who, what, how, (and why) breakdown. Who == radar charts, how == stacked area, what == tree and why == historical context (i.e. Timeline).
nathaniel
2016-04-25 18:58
okay ? I?m going get through as much as I can in the next hour. Then I?m offline until about 9:30 pm tonight unfortunately (I may finish up a bit sooner).
jerclark
2016-04-25 18:58
OK.
jerclark
2016-04-25 18:58
I can't really tell if we're going to get graded on Prototype V2...
jerclark
2016-04-25 18:59
It looks like the point allocation is 0...
jerclark
2016-04-25 18:59
We got a little roughed up on V1 :wink:
nathaniel
2016-04-25 18:59
Yeah, I can?t either. I mean I suspect that we?re *actually* going to be graded on the final version and they are just trying to push us along, but who knows.
jerclark
2016-04-25 18:59
I hope so!
nathaniel
2016-04-25 19:00
I don?t think they have ever given us a brake down of how much each part of the project is worth.
jerclark
2016-04-25 19:00
But I do feel pretty good about the baseline we've got, even though the story piece isn't in place yet. It's an ambitious technical effort and we've got most of it working.
nathaniel
2016-04-25 19:01
me too
jerclark
2016-04-25 22:15
Bruno, you gonna be online tonight?
bruno-c
2016-04-25 23:41
yup
jerclark
2016-04-25 23:42
nice.
bruno-c
2016-04-25 23:47
for the story bit, i was thinking that it would be cool to essentially have "preset" stories
bruno-c
2016-04-25 23:47
so let's say on the right of the radar or whatever
jerclark
2016-04-25 23:47
what?d you have in mind?
bruno-c
2016-04-25 23:49
there's a "carousel" of say 3 things [1, 2, 3, done] story 1 say is about housing at a specific point in time so when you select story one, the year filters and category / demographic / items selectors switch by themselves
bruno-c
2016-04-25 23:49
and then the story sortof narrates the data. and then when you're done you can skip to the next story, or just freeform
jerclark
2016-04-25 23:49
I get it - like example stories, that show how to create new ones.
jerclark
2016-04-25 23:49
I like that.
jerclark
2016-04-25 23:50
Can you take a crack at that? Feel free to adjust sizes of the visualizations...
jerclark
2016-04-25 23:50
I?m cleaning up the tree thingy a bit, I?ll ping you when I can help with story/layout.
bruno-c
2016-04-25 23:52
cool yeah i'll give it a go. if ideas for stories come to you let me know. i wrote some down and now i can't find the paper i wrote it down on haha. probably my cat
bruno-c
2016-04-26 00:15
hmm note to self: when we calculate the percentages, maybe we should use "percentage of how much people have spent"
jerclark
2016-04-26 00:16
Is that what it is?
bruno-c
2016-04-26 00:16
right now it's a percentage of before tax income
bruno-c
2016-04-26 00:17
but i'm thinking like, maybe that other number could be more relevant, and that seems to be how the bls itself does that sort of calculation
bruno-c
2016-04-26 00:17
bruno-c
2016-04-26 00:18
incidentally following links from there and generally the "beyond the numbers" thing is a treasure trove of little stories
bruno-c
2016-04-26 00:18
i think i'm just gonna go right ahead and paraphrase those and cite them as reference. it's not like i'm gonna come with relevant original ones
bruno-c
2016-04-26 00:19
i'm gonna add a lot more help and definition of terms as well like a glossary
jerclark
2016-04-26 00:19
I think that?s a good idea.
jerclark
2016-04-26 00:20
Although I like the % of income for some reason...
jerclark
2016-04-26 00:20
What happened to delta between income earned and expenditures?
bruno-c
2016-04-26 00:20
savings i guess?
bruno-c
2016-04-26 00:20
put it in a sock
jerclark
2016-04-26 00:21
Although the % of expenditures would make the small expenditures possibly look bigger.
jerclark
2016-04-26 00:21
You got a minute to chat?
2016-04-26 00:22
@jerclark has started a Google+ Hangout for this channel. https://plus.google.com/hangouts/_/leucdauq7vhb7pguflo4mskgpye.
bruno-c
2016-04-26 00:25
gimme 5 i'll be there
jerclark
2016-04-26 00:27
Sure
jerclark
2016-04-26 01:12
@jerclark uploaded a file: https://melodictransport.slack.com/files/jerclark/F13LEUV7A/-.js
jerclark
2016-04-26 01:12
@jerclark uploaded a file: https://melodictransport.slack.com/files/jerclark/F13K13YSX/-.js
bruno-c
2016-04-26 02:58
added basic presidents event markers to timeline
bruno-c
2016-04-26 02:58
gonna try to make the ui less gross
jerclark
2016-04-26 03:47
Cool.
jerclark
2016-04-26 03:47
Let me know when you?re done.
jerclark
2016-04-26 03:48
I?ve got some changes that need to be merged?we should probably get started on that.
jerclark
2016-04-26 03:49
Bruno are you working on a branch?
jerclark
2016-04-26 03:51
I have to hand something in in just a couple minutes
jerclark
2016-04-26 03:51
nathaniel, do you want to merge?
nathaniel
2016-04-26 03:53
Sure
jerclark
2016-04-26 03:53
nothing like cutting it close
nathaniel
2016-04-26 03:53
Yeah
jerclark
2016-04-26 03:53
It looks like your branch won?t automatically merge...
jerclark
2016-04-26 03:54
You there bruno?
nathaniel
2016-04-26 03:55
I?m running into conflicts when I try to merge master into my branch
nathaniel
2016-04-26 03:55
I would just leave it out for now
nathaniel
2016-04-26 03:55
The changes are very miner
jerclark
2016-04-26 03:55
OK.
jerclark
2016-04-26 03:55
I?m just going to hand in what?s in master.
nathaniel
2016-04-26 03:55
Yes
nathaniel
2016-04-26 03:55
that makes sense
bruno-c
2016-04-26 03:56
ack
bruno-c
2016-04-26 03:56
uyeah
bruno-c
2016-04-26 03:56
i put a bunch of stuff in there
bruno-c
2016-04-26 03:56
:simple_smile:
bruno-c
2016-04-26 03:56
JUST pushed
bruno-c
2016-04-26 03:57
everything looks like it's still working after i put the story component
bruno-c
2016-04-26 03:58
aw fuck i had muted the notif
jerclark
2016-04-26 04:01
I had to submit Let me see if I can submit again...
bruno-c
2016-04-26 04:02
it's such a huge file too
jerclark
2016-04-26 04:05
I?ll try again, see if they take it,
jerclark
2016-04-26 04:05
I like the story picker.
jerclark
2016-04-26 04:08
It allowed me to submit it?though it has omnious red letters saying ?LATE? :wink:
bruno-c
2016-04-26 04:09
well, let's pray. if i recall correctly this is the class that was all like if you're late you're dead
jerclark
2016-04-26 04:10
@jerclark uploaded a file: https://melodictransport.slack.com/files/jerclark/F13LYHRHA/pasted_image_at_2016_04_26_12_10_am.png
bruno-c
2016-04-26 04:10
hopefully they won't be total dicks
bruno-c
2016-04-26 04:10
in the parlance of our time
jerclark
2016-04-26 04:10
Well, if it?s any consolation, I did hand something in at 11:59. That should count for something. I didn?t think it would completely obliterate my original submission.
jerclark
2016-04-26 04:11
Thought it might be like a ?history??that way, we could at least get graded on the original submission if they didn?t want to take the late one.
bruno-c
2016-04-26 04:11
yeah i think the whole point of vocareum is that it keeps a history
bruno-c
2016-04-26 04:11
yeah the story thing? i mean it's really really really rough
jerclark
2016-04-26 04:11
I think it?s a good idea though.
bruno-c
2016-04-26 04:11
but generally that's the idea. like when you click the second story you can see the data changes
bruno-c
2016-04-26 04:11
because the item changed
jerclark
2016-04-26 04:12
I think it?s a good idea?Tell a couple simple stories, then let the users loose.
bruno-c
2016-04-26 04:12
yeah. i think we're gonna need a much better demographic / category / item picker of some kind
jerclark
2016-04-26 04:12
I think it should all be at the top in a single control bar.
bruno-c
2016-04-26 04:13
yeah maybe collapsible
jerclark
2016-04-26 04:13
If we put it up there, you can just call wrangleData on my components and I?ll read from those menus.
bruno-c
2016-04-26 04:14
yeah we'll figure that out. i wonder how much the code is worth in terms of points
bruno-c
2016-04-26 04:14
i means it's CS 171, not design 171
bruno-c
2016-04-26 04:14
not a whole lot of CS has happened if any
bruno-c
2016-04-26 04:14
oh well ¯\_(?)_/¯
jerclark
2016-04-26 04:14
That?s true.
jerclark
2016-04-26 04:15
My code is gettin? pretty sloshy at this point...
bruno-c
2016-04-26 04:15
yeah so is mine, i did some very bad things to get things to work when i had 2 minutes left lol
bruno-c
2016-04-26 04:16
after i finish that hellish problem set for cs51 i'll do some code cleanup that's something that i'm really good at. like a code janitor
bruno-c
2016-04-26 04:16
architect / code janitor, that's my title
jerclark
2016-04-26 04:16
Sweet. Alright - I?m signing off for now.
jerclark
2016-04-26 04:17
Catch up in a day or two.
bruno-c
2016-04-26 04:17
cool, have a good one!
jerclark
2016-04-28 01:19
Yo dudes. You planning to do anything in the coming days?
nathaniel
2016-04-28 01:21
Yes! But I?m now basically occupied until Friday evening?
jerclark
2016-04-28 01:31
OK. that?s cool. I?ll keep you posted of anything I do.
jerclark
2016-04-28 19:41
Did we have to do something for a presentation on Tuesday?
jerclark
2016-04-28 20:22
Just saw on piazza that we didn't...
bruno-c
2016-04-28 23:54
Yup, but we will need another poster apparently
bruno-c
2016-04-29 00:02
If you want to feel good and / or bad, have a look at previous years winners http://laurenwood.github.io/
bruno-c
2016-04-29 00:02
oops
jerclark
2016-04-29 02:13
Thanks for sharing?yeah, not sure if I feel good or bad. But I do think what we have is somewhat interesting.
jerclark
2016-04-30 00:18
You guys gonna be online tonight. Divide and conquer.
bruno-c
2016-04-30 02:01
Whole weekend is booked for this. :simple_smile:
bruno-c
2016-04-30 02:02
i really like what you did with the labels on the radar chart. It looks strangely.. occult
bruno-c
2016-04-30 02:11
I?ll be online tomorrow morning early. I?ll catch you guys sometimes tomorrow. I?m going to focus on stories and layout and generally making things look pretty and flowing
jerclark
2016-04-30 02:20
That?s awesome?I was gonna break ground on that, but figured I?d wait too see if you wanted to take a crack at it.
jerclark
2016-04-30 02:23
One story I found is ?Bringing home the bacon (just less of it, that?s all)"
jerclark
2016-04-30 02:24
Pork by Housing type, 2004-2014. Basically, uniform distribution of pork consumption by all housing types, and uniform reduction in consumption between 2004-2014:
jerclark
2016-04-30 02:24
@jerclark uploaded a file: https://melodictransport.slack.com/files/jerclark/F153DGP8Q/pasted_image_at_2016_04_29_10_24_pm.png
jerclark
2016-04-30 02:24
Health reasons?
jerclark
2016-04-30 14:19
Bruno, I looked into CU data to get CPI values that should work for the tree chart. You got a second to chat?
bruno-c
2016-04-30 14:57
yup
2016-04-30 14:59
@jerclark has started a Google+ Hangout for this channel. https://plus.google.com/hangouts/_/rkm5n5jtfjdgdbctydbkfysekme.
bruno-c
2016-04-30 16:19
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F154S7NE8/Untitled.html
bruno-c
2016-04-30 16:19
@bruno-c pinned their HTML https://melodictransport.slack.com/files/bruno-c/F154S7NE8/Untitled.html to this channel.
nathaniel
2016-04-30 16:49
Hi Guys ? I?m going to be a few minutes late to our scheduled 1 pm chat. Can you both do 1:15?
bruno-c
2016-04-30 16:49
fine with me
nathaniel
2016-04-30 16:49
Great ?thanks.
bruno-c
2016-04-30 16:51
@jerclark: i have to go play in the tree.js file to fix some layout issues, if you have some updates you want to push to master go ahead
nathaniel
2016-04-30 17:20
Or, are you guys in the other hangout?
bruno-c
2016-04-30 17:21
no no, joining now
jerclark
2016-04-30 17:42
Did you guys already sign off?
nathaniel
2016-04-30 17:42
We did
jerclark
2016-04-30 17:42
All good?
nathaniel
2016-04-30 17:42
but happy to jump back on if you want
nathaniel
2016-04-30 17:42
I think we?re good for now
jerclark
2016-04-30 17:42
OK.
bruno-c
2016-04-30 18:29
couple of pointers for the layout
bruno-c
2016-04-30 18:29
i switch to container instead of container-fluid. usually, this should make no difference to you
bruno-c
2016-04-30 18:30
if you have to do some grid layouts, remember you only need one `.container` div
bruno-c
2016-04-30 18:30
after that, you keep nesting `.row` but you don?t have to use container again
bruno-c
2016-04-30 18:31
e.g ` .container > .row > .col-sm-6 > .row > .col-sm-6, .col sm-6` etc
jerclark
2016-04-30 18:31
OK. The Multi-plot template uses grid, and sets a container. We can remove that if the root element is a container.
bruno-c
2016-04-30 18:31
everytime you nest, you have 12 units to use again
bruno-c
2016-04-30 18:31
yeah i?ve fixed all that stuff now :simple_smile:
jerclark
2016-04-30 18:31
Cool.
jerclark
2016-04-30 18:31
Thx.
bruno-c
2016-04-30 18:31
that why i was saying i?ll have to go play in the tree.js
bruno-c
2016-04-30 18:32
oh and use `.col-sm-x` instead of `.col-md-x`
bruno-c
2016-04-30 18:33
it should prevent the layout from breaking at smaller widths, it will become non-responsive
jerclark
2016-04-30 18:33
Ah - I misses that. No updates yet. Just got back on about 30 min ago, just finishing up CPI data wrangling. Also moving rename to the cleanup script (by virtue of an alternate ?cx.characteristics? file with renamed names.
bruno-c
2016-04-30 18:33
good idea.. i was worried we?d overwrite your changes
jerclark
2016-04-30 18:33
Did you make the switch from md to sm?
bruno-c
2016-04-30 18:33
yah
jerclark
2016-04-30 18:33
OK.
jerclark
2016-04-30 18:33
Just checking.
bruno-c
2016-04-30 18:33
all good
bruno-c
2016-04-30 18:35
bruno-c
2016-04-30 18:35
need some nice background for the old people stories hehe
bruno-c
2016-04-30 19:46
@jerclark: when we talked earlier you had suggested that the radar was good for what again? why / what / how ? i forget :simple_smile:
jerclark
2016-04-30 19:47
who.
jerclark
2016-04-30 19:48
Whose spending more on entertainment? Elderly or young? Northeasterners or Westerners? Low income earners or High-income earners?
jerclark
2016-04-30 19:49
Who eats the most eggs? Buys the most personal care items?
bruno-c
2016-04-30 20:39
@jerclark: i should have taken notes. Area Chart would be: What. Money Tree: How?
jerclark
2016-04-30 20:41
I think that could be spun either way.
jerclark
2016-04-30 20:41
I had it the other way around...
jerclark
2016-04-30 20:42
I think maybe the way you mention above is better.
jerclark
2016-04-30 20:42
Area == what, tree == how
bruno-c
2016-04-30 20:42
copy is like
bruno-c
2016-04-30 20:42
> The <a href="#radar">Radar Chart</a> is useful for exploring <em>who</em> is spending on what. Whose spending more on entertainment? Elderly or young? Northeasterners or Westerners? Low income earners or High-income earners? Who eats the most eggs? Buys the most personal care items?</li>
jerclark
2016-04-30 20:43
Whose should be ?Who is? (I know, that?s my copy?)
bruno-c
2016-04-30 20:43
lol
bruno-c
2016-04-30 20:43
> <p>The <a href="#area">Area Chart</a> lets you drill down into the details of <em>what</em> the average american spends their money on. Expenditure items are grouped in convenient categories that can be expanded to examine the details.</p>
bruno-c
2016-04-30 20:44
whose on first
bruno-c
2016-04-30 20:44
...
jerclark
2016-04-30 20:44
And maybe: ?is useful for exploring ?who? is consuming what"
bruno-c
2016-04-30 20:44
got it
jerclark
2016-04-30 20:44
I think the area text is good.
jerclark
2016-04-30 20:45
The tree gives insight into ?how? the average american allocates key parts of their budget.
bruno-c
2016-04-30 20:48
<p>The <a href="#money-tree">Money Tree</a> is a unique visualization that gives insight into <em>how</em> the average american allocates key parts of their budget. In particular, it highlights the relative weight of those expenditures and the effect of changes in buying power.<li>
bruno-c
2016-04-30 20:49
it?s a start. layout is starting to come together
bruno-c
2016-04-30 21:23
ungh sorry i added an 8M large jpg. lol
bruno-c
2016-04-30 21:23
didn?t notice the size. it?s not in use in the page
jerclark
2016-04-30 21:39
Money Tree looks good.
jerclark
2016-04-30 21:40
Like it.
bruno-c
2016-04-30 22:42
@jerclark: would it make sense to dynamically pick the expenditures category in the money tree so that you don?t get ?0%? item basically? like, pick the top 14
bruno-c
2016-04-30 22:46
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F155PSECU/consumer-47205.png
jerclark
2016-04-30 22:51
Maybe
jerclark
2016-04-30 22:51
:wink:
jerclark
2016-04-30 22:52
We could filter the data, sure.
jerclark
2016-04-30 22:52
Just including everything for now
jerclark
2016-04-30 22:52
I?m working on getting the scaling right for the CPI stuff, then that should be all set. Data is loading and all.
bruno-c
2016-04-30 22:53
great!
bruno-c
2016-04-30 22:53
i just pushed a version that has a background for the trees, as an experiment. let me know what you think
bruno-c
2016-04-30 22:53
would have to tweak the trees a bit but i think it really ties the page layout together
jerclark
2016-04-30 22:54
The scale of the icon will be based on the difference between the subcategory-specific CPI and the ?AllItems? cpi (the closest thing I can find to the ?overall inflation rate). So, for example, if alcohol has a lower CPI than AllItems, it will appear ?bigger?. You can get more alcohol for your money.
jerclark
2016-04-30 22:54
Compared to the average of all items.
jerclark
2016-04-30 22:54
Let me take a look.
bruno-c
2016-04-30 22:54
awesome. i was just thinking about the icons wondering if we could use scale on them
jerclark
2016-04-30 22:54
Should I pull from master?
bruno-c
2016-04-30 22:55
yes, hopefully it won?t be too painful a merge
bruno-c
2016-04-30 22:55
i can do the integration too if you want, whatever you prefer
jerclark
2016-04-30 22:55
I?ll commit my changes to a branch, stash, then checkout master just to check it out. I?ll integrate in a bit. Shouldn?t be to hard.
bruno-c
2016-04-30 22:55
cool
jerclark
2016-04-30 23:01
I definitely like the background concept! We could tweak the trees no problem. Feel free to play around with the styles in the tree.css file...
jerclark
2016-04-30 23:02
I have to go to a gathering for a little while, I?ll either be online a bit later or pick it up early tomorrow.
jerclark
2016-04-30 23:02
But I?m feeling pretty good about the CPI integration. That was definitely the biggest outstanding thing for me, and I think it will add to the tree view.
bruno-c
2016-04-30 23:02
cool, talk to you later. if you?re in a position to push your changes that?d be great
jerclark
2016-04-30 23:04
I can check in, but the scaling is bit off. Hold on.
jerclark
2016-04-30 23:10
It?s all merged to master.
bruno-c
2016-04-30 23:10
awesome!
jerclark
2016-04-30 23:11
I also replicated the incomes and incomeForYear functions to provide similar functionality for CPI, but had to drop the memoize cause the hash function was causing a single value to cache. I?ll need to revisit that if we need to optimize.
jerclark
2016-04-30 23:12
Feel free to poke around with the scaling on the tree ?leaves??I was trying to think of the right function to come up with an SVG scale value. SVG scale values are based on ?1?.
jerclark
2016-04-30 23:12
0.5 shrinks by half, 2 doubles in size.
jerclark
2016-04-30 23:13
I?ll probably jump back on in a bit. Layout is looking good, btw.
bruno-c
2016-04-30 23:14
thanks. lots left to do but i think we?re doing good
jerclark
2016-05-01 14:44
Either of you guys online?
bruno-c
2016-05-01 15:03
i?m on
jerclark
2016-05-01 15:06
Cool.
jerclark
2016-05-01 15:06
How you feeling today? Ready to kill it?
bruno-c
2016-05-01 15:07
yeh for sure. last homework due for this session so
jerclark
2016-05-01 15:07
Oui.
jerclark
2016-05-01 15:07
On y va.
bruno-c
2016-05-01 15:08
:+1:
bruno-c
2016-05-01 15:10
what do you think of the shopping lady
bruno-c
2016-05-01 15:11
i?m gonna make the background white i think, try to replicate some of the design ideas from my midterm project to a certain extent
jerclark
2016-05-01 15:11
Je l?aime. Qu?est que tu pense des colours du ?radar-chart?? Rouge et bleu sont merdique
bruno-c
2016-05-01 15:11
Oui, je suis daccord.
bruno-c
2016-05-01 15:12
+ confusing because of the timeline
jerclark
2016-05-01 15:12
Ouai. Est q?Il y a une thème de couleur?
bruno-c
2016-05-01 15:14
Oui :simple_smile: Dans `styles.css` it says ?palette"
bruno-c
2016-05-01 15:14
but i may change it
jerclark
2016-05-01 15:23
I think white background, with interstitial block quotes is a good play here.
bruno-c
2016-05-01 15:23
yeah
jerclark
2016-05-01 15:24
Will require a slightly better narrative than we?ve got now, but the who-what-how think is a good structure.
jerclark
2016-05-01 15:24
Alright, I?m gonna keep working.
jerclark
2016-05-01 15:24
I cleaned up that tip leak, btw.
jerclark
2016-05-01 15:31
Any thoughts on why the radar hovers are so flaky?
bruno-c
2016-05-01 15:32
yeah the elements are often really close together
bruno-c
2016-05-01 15:32
there?s some trick with pointer-events, lemme check
bruno-c
2016-05-01 15:34
hmm, all i can find is from an area chart i made at one point
bruno-c
2016-05-01 15:35
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F1568NZPD/-.js
bruno-c
2016-05-01 15:36
but i don?t remember what i was doing lol. :simple_smile: it would be cool if the points could be ?selected? by clicking on them so that when you click the tip remains displayed until you click another node
bruno-c
2016-05-01 15:36
but that can be an enhancement if we have time
jerclark
2016-05-01 16:47
Natural Disasters:
jerclark
2016-05-01 16:47
Loma Prieta - california earthquate
jerclark
2016-05-01 16:47
Katrina
jerclark
2016-05-01 16:48
Superstorm Sandy
jerclark
2016-05-01 16:48
California drought
jerclark
2016-05-01 16:48
Wars:
jerclark
2016-05-01 16:48
Gulf War 1
jerclark
2016-05-01 16:49
Gulf War 2
jerclark
2016-05-01 16:49
9/11
jerclark
2016-05-01 16:50
End of cold war (Berlin Wall)
jerclark
2016-05-01 16:50
Financial Events
jerclark
2016-05-01 16:51
2008 (housing crash and great recession)
jerclark
2016-05-01 16:51
2001 (dot-com bubble birst)
jerclark
2016-05-01 16:51
1995-2001 (dot-com boom - bull market)
bruno-c
2016-05-01 19:02
@jerclark: alright i pushed a timeline with multiple event types
bruno-c
2016-05-01 19:03
still need to figure out how to display the associated information
bruno-c
2016-05-01 19:03
(labels overlap)
jerclark
2016-05-01 19:06
I think it would be OK to only show 1 at a time.
jerclark
2016-05-01 19:06
(select dropdown style)
bruno-c
2016-05-01 19:06
yeah, that?s how it is now
bruno-c
2016-05-01 19:06
but since there?s only 4 choices of timelines, it fits nicely as a ?navigation"
jerclark
2016-05-01 19:15
So you mean overlapping labels within one category. Got it.
bruno-c
2016-05-01 19:36
I think i?m gonna have to go with more tips.
jerclark
2016-05-01 19:38
Bruno, I just checked in some tree updates. If item is tracking slower than inflation (good for consumers), there?s a pink ?flower?. If item is tracking faster than inflation, the leaf starts to turn brown. Have a look. Also made a little mulch bed around the base of the tree.
jerclark
2016-05-01 19:38
Moving on from that for a while.
jerclark
2016-05-01 19:40
I feel like we need to move the text and controls to the side of the trees (and potentially make them smaller), so we don?t have so much empty space above the trees.
bruno-c
2016-05-01 19:45
i did a tiny little tweak to the tree now to allow the branches to ?pop out? of their containing elements"
bruno-c
2016-05-01 19:45
looks really good!
bruno-c
2016-05-01 19:47
one thing i?m wondering is, how do you calculate the height of the tree? say I take year 1984, would it be relative to the average salary across the board?
bruno-c
2016-05-01 19:49
i?ll work on the layout around the trees now that they are close to their final forms
bruno-c
2016-05-01 19:49
i like using the entire horizontal space because they look great
jerclark
2016-05-01 19:53
Good question. The height of the tree is absolute, pegged to 2014 adjusted dollars. The domain is $5000 - $100000, and the range is 0 to full height of svg. Height is capped so that anything over $100000 displays at full height of SVG (which is a little weird, but I wasn?t sure how to allow for unlimited growth).
jerclark
2016-05-01 19:53
We could do something relative or log based, but I think the linear scale allows for differences to really pop.
jerclark
2016-05-01 19:53
That?s 5K to 100K in the domain.
jerclark
2016-05-01 19:54
I?ve discovered an issue with the data that will need fixing, but it shouldn?t require any layout changes.
bruno-c
2016-05-01 19:56
i?m just wondering if it would be more accurate if you used the range of revenues for the year you?re looking at. like if you?re looking at < 25yr olds in 1984 you?d be in much better shape making 25k then you would be in 2014
bruno-c
2016-05-01 19:56
although that shows in the branches too come to think of it
bruno-c
2016-05-01 20:20
i?m gonna get something to eat and i?ll be back to solve the timeline label issue
bruno-c
2016-05-01 20:20
have an idea
jerclark
2016-05-01 20:42
Bruno, is there a way we can switch between ?live? updates and ?buffered? updates in the timeline? It looks like you switched it to only reload after the brush is over?but we lose the ?animation? factor in the radar and tree charts?which was a really cool way to observe change over time.
bruno-c
2016-05-01 20:43
it was too jerky. we could change it to do throttling instead of ?debouncing? maybe and see what happens
bruno-c
2016-05-01 20:44
it?ll have a try
jerclark
2016-05-01 20:45
And here's a label possibility:
jerclark
2016-05-01 20:45
@jerclark uploaded a file: https://melodictransport.slack.com/files/jerclark/F1588S1PG/slack_for_ios_upload.jpg
bruno-c
2016-05-01 20:47
ah very cool. yeah i could try that!
bruno-c
2016-05-01 20:47
might make it a little taller or maybe we?ll just tweak the data until it fits nicely
bruno-c
2016-05-01 20:48
it would make it possible to show date ranges as well actually
bruno-c
2016-05-01 20:59
@jerclark: was the animation ever working smoothly for you? i?m trying with different parameters here and at best it?s super jerky, like 1 frame per second
bruno-c
2016-05-01 21:00
maybe svg drawing is slower on my computer, could be that
bruno-c
2016-05-01 21:15
great idea, i did the label thing
bruno-c
2016-05-01 21:16
i disabled the throttling everywhere, maybe have a look if it does what you want
jerclark
2016-05-01 21:24
That is the desired behavior - but it is kinda slow. Maybe we could offer a ?switch? that says ?animate charts while dragging? to disable the throttling.
jerclark
2016-05-01 21:25
This will imply that it might slow down, or we could say something like (will decrease performance)
jerclark
2016-05-01 21:25
That way, people can choose if they want to ?animate? or not.
jerclark
2016-05-01 21:25
Its also possible that some of the performance could be improved if we memoize the cpi methods on the dataset?
bruno-c
2016-05-01 21:25
i put a lot of tracing for the speed in the code, so it?s pretty easy to know where the time is being spent
bruno-c
2016-05-01 21:26
definitely if you have to massage that data often (i haven?t a chance to look) it would help
bruno-c
2016-05-01 21:26
or structure it differently if it?s not by index already
jerclark
2016-05-01 21:26
It could use some attention. Maybe we can chat in a little bit.
bruno-c
2016-05-01 21:27
okay. let?s take a second to recap again now if you don?t mind
bruno-c
2016-05-01 21:28
what i want to do next: wrap up the design and layout around the money tree
bruno-c
2016-05-01 21:28
definitely want that part to look really good
bruno-c
2016-05-01 21:29
then do a bit of tweaking for the design on the radar
bruno-c
2016-05-01 21:29
then figure out what is happening with the cake
bruno-c
2016-05-01 21:29
then stories
bruno-c
2016-05-01 21:30
and then logistics and cleanup - the video, cleaning up the repo and code before handing it in, all that stuff
bruno-c
2016-05-01 21:30
and the poster (ungh) what is up with posters in this class
jerclark
2016-05-01 21:39
Cool. I?ve hit a snag with the CPI data for education, which I think is an important category. It?ll take me a little longer to figure it out.
jerclark
2016-05-01 22:05
I?ve resolved the data issue - but it will need a footnote!
jerclark
2016-05-01 22:06
As for the slowness - it?s definitely my cpi data code. I wrapped the functions in _.memoize and it?s much faster now, but now I get the same values back when I don?t want them?have to take a look.
jerclark
2016-05-01 22:21
Bruno - you got 5 to chat?
jerclark
2016-05-01 22:21
Need some help with memoization.
bruno-c
2016-05-01 22:22
Ok cool i have some time, dinner is cooking
jerclark
2016-05-01 22:22
Sure.
2016-05-01 22:23
@jerclark has started a Google+ Hangout for this channel. https://plus.google.com/hangouts/_/yk6airjfgvdjpi3ltmi5krcabae.
bruno-c
2016-05-01 22:30
@bruno-c uploaded a file: https://melodictransport.slack.com/files/bruno-c/F158VEZDF/-.js
bruno-c
2016-05-01 22:40
eating dinner now. @nathaniel if something is making the layout break wider than your monitor maybe check with debugging console to figure out which element might be at fault
bruno-c
2016-05-01 22:40
i changed some stuff with the svg elements that allows them to overflow outside their containers, maybe it has to do with that
jerclark
2016-05-02 00:40
@jerclark uploaded a file: https://melodictransport.slack.com/files/jerclark/F159C6VQA/-.html
jerclark
2016-05-02 00:41
I added a possible story for the tree section as a snippet.
jerclark
2016-05-02 00:41
Bruno, do you want me to set those up, or is giving you copy like this OK?
bruno-c
2016-05-02 00:42
this is fine
bruno-c
2016-05-02 00:46
i just merged some changes from the area chart branch @nathaniel
nathaniel
2016-05-02 00:46
k
bruno-c
2016-05-02 00:52
crazy idea - how about we do multiple small area charts
nathaniel
2016-05-02 00:53
What would each chart show?
bruno-c
2016-05-02 00:53
each category you have at the bottom?
bruno-c
2016-05-02 00:53
i?m finding it really hard to find a way to have it go with the document flow kinda
bruno-c
2016-05-02 00:54
i guess when i made it giant in the first place i didn?t figure it might remain that way :stuck_out_tongue:
jerclark
2016-05-02 00:54
@jerclark uploaded a file: https://melodictransport.slack.com/files/jerclark/F157LQT19/housing_halo.html
nathaniel
2016-05-02 00:54
Small charts is going to be hard if you want any kind of labeling of the items (but not impossible)
bruno-c
2016-05-02 00:55
It?s really cool what you did with the zooming it and the broken line and labels though
nathaniel
2016-05-02 00:55
The idea I always wanted to try was two side-by-side area charts
nathaniel
2016-05-02 00:55
But right now I?m just trying to get this clean and nice
bruno-c
2016-05-02 00:57
okay when we get to that point we may want to try something
bruno-c
2016-05-02 01:18
@jerclark: think we still need the year label at the top of the tree or the bottom one is enough?
bruno-c
2016-05-02 01:18
i feel like one is probably fine
jerclark
2016-05-02 01:21
Bottom is probably better?
bruno-c
2016-05-02 01:22
k got it
jerclark
2016-05-02 01:22
I?ll drop the top.
bruno-c
2016-05-02 01:22
i?m on it
jerclark
2016-05-02 01:22
OK.
jerclark
2016-05-02 01:23
Is there anything either of you want me to work on right now?
jerclark
2016-05-02 01:23
I can do some more cleanup/enhancement, but I?m also open to get other stuff just working.
bruno-c
2016-05-02 01:23
if you still have more good stories :simple_smile: also the copy
bruno-c
2016-05-02 01:23
like the introduction paragraphs and all that, if you think it needs editing
jerclark
2016-05-02 01:24
OK.
jerclark
2016-05-02 01:24
I was thinking 2 stories per section would be good, eh?
bruno-c
2016-05-02 01:24
yup that?s good
jerclark
2016-05-02 01:28
I really like the intro section for the most part. I can make a few edits.
jerclark
2016-05-02 01:40
I really do like the copy. I will review later and see if it could use some changes.
jerclark
2016-05-02 02:12
I actually made some copy updates, and created a PR. You can merge them if you want to.
jerclark
2016-05-02 02:13
Much nicer colors on the radar! I think we need to put the demographic pickers closer to the trees. They get hidden where are now.
jerclark
2016-05-02 02:15
Like:
jerclark
2016-05-02 02:15
@jerclark uploaded a file: https://melodictransport.slack.com/files/jerclark/F157V0Y2X/slack_for_ios_upload.jpg
jerclark
2016-05-02 02:19
Are you guys still chugging along?
bruno-c
2016-05-02 02:21
yeah for a bit
bruno-c
2016-05-02 02:21
i?m starting to get stupid
jerclark
2016-05-02 02:22
Me too.
bruno-c
2016-05-02 02:22
the problem with that for me is that because the height of the tree varies a lot, you get these dropdowns that float in the air above the trees :simple_smile:
jerclark
2016-05-02 02:23
Fair point. Can we put them down below next to the trees?
jerclark
2016-05-02 02:23
And tuck the trees up below the copy?
bruno-c
2016-05-02 02:23
switch to highest 20% :simple_smile:
bruno-c
2016-05-02 02:24
we could have the tree go over the containers though, that could work
jerclark
2016-05-02 02:24
Tucked up already...
jerclark
2016-05-02 02:25
We could make the trees smaller.
jerclark
2016-05-02 02:25
I?m not really sure how to handle the ?white space? above the trees!
bruno-c
2016-05-02 02:25
but they?re so pretty!
bruno-c
2016-05-02 02:25
well, we can position them absolutely
jerclark
2016-05-02 02:26
How about controls in between the trees? There should be room, right?
jerclark
2016-05-02 02:27
Like halfway up the trees, in between them.
bruno-c
2016-05-02 02:28
i can maybe affix them to the document, same way the timeline works
bruno-c
2016-05-02 02:29
so when you scroll past them they stay attached to the left side of the timeline
bruno-c
2016-05-02 02:29
the help text too is so long, maybe it could show / hide
bruno-c
2016-05-02 02:37
k i pushed that to demonstrate. it might work, or something similar
bruno-c
2016-05-02 02:37
i think i?m gonna call it a night. just gonna have to make some time in the morning to wrap up remaining things
jerclark
2016-05-02 02:47
I checked something in to ?absolute-pickers-experiment??I stuck em between the trees...
bruno-c
2016-05-02 02:52
if you?re still around check out the latest thing i put in
bruno-c
2016-05-02 02:52
just a splash of color but it really brings everything together, especially the trees. builds up to it :simple_smile:
jerclark
2016-05-02 03:22
yes?I like the blue background...
jerclark
2016-05-02 03:22
We?ll have to wrangle the area chart - beyond that. I think it?s looking pretty good.
jerclark
2016-05-02 03:46
Nathaniel - one thing for area chart?are you planning to allow click through to single item?
nathaniel
2016-05-02 22:33
Hi Guys ? sorry for taking so long on this, but the area-chart stuff is now complete
nathaniel
2016-05-02 22:34
you can zoom in all the way to a single layer and the transitions are at least, not completely ridiculous.
nathaniel
2016-05-02 22:34
(I have not merged into master)
jerclark
2016-05-02 22:44
Sweet! Looking forward to taking a look.
jerclark
2016-05-02 22:44
Do you guys want to get online tonight? Maybe an active working session to finish up?
nathaniel
2016-05-02 22:45
Yes ? I was about to go for a quick run ? do you want to say 7:30 or 8 pm?
nathaniel
2016-05-02 22:45
(but I?m free whenever)
jerclark
2016-05-02 22:46
Sure. Ping when you get back
jerclark
2016-05-02 22:58
You available bruno?
bruno-c
2016-05-02 23:03
Oh yeah I'm just finding some dinner and then one last sprint.
jerclark
2016-05-02 23:05
I merged nathaniels changes into master.
jerclark
2016-05-02 23:34
Bruno - how do you set the default story?
bruno-c
2016-05-02 23:35
It chooses the first one it finds in the markup
jerclark
2016-05-02 23:36
Cool?I?m seeing some bugs. Maybe we can work together to hammer them out?
bruno-c
2016-05-02 23:40
For sure. I'm just finishing dinner I'll be right there.
nathaniel
2016-05-02 23:46
Back from my run
nathaniel
2016-05-02 23:46
I?ll be back online in about 10 min
bruno-c
2016-05-02 23:51
oh oh hmm. yeah if we have multiple stories in different sections, that would mean that only one can be selected
bruno-c
2016-05-02 23:51
otherwise the timeline will be wrong
bruno-c
2016-05-02 23:58
found the radar title problem
nathaniel
2016-05-03 00:03
back
jerclark
2016-05-03 00:03
OK.
jerclark
2016-05-03 00:03
Cool.
nathaniel
2016-05-03 00:03
do you guys want to hop on a google hangout?
jerclark
2016-05-03 00:03
Yes. google hangout.
bruno-c
2016-05-03 00:03
yup
nathaniel
2016-05-03 00:04
let me find our last link..
2016-05-03 00:05
@jerclark has started a Google+ Hangout for this channel. https://plus.google.com/hangouts/_/u25uaexi5vcb3ggzgqlfy7h2wue.
jerclark
2016-05-03 00:05
Nathanie - you can do ?/hangout? in the slack channel and it sets one up.
nathaniel
2016-05-03 00:05
did not know that!
nathaniel
2016-05-03 00:05
useful
jerclark
2016-05-03 01:05
Process book:
jerclark
2016-05-03 01:06
@jerclark uploaded a file: https://melodictransport.slack.com/files/jerclark/F15HV2PCL/pasted_image_at_2016_05_02_09_06_pm.png
nathaniel
2016-05-03 01:06
Nice!
bruno-c
2016-05-03 01:06
:simple_smile: