search

arpitg1991 / stop-frisk-analysis

Data analysis on nyclu stop and frisk data
1 stars 3 forks source link

working with d3 #1

Open bchnge opened 11 years ago

bchnge commented 11 years ago

Hi all,

Just letting you know that i've set up a working example of the NYC map in d3 in index.html. Next steps are to incorporate external data files and implement interactive features (e.g. tooltips, zooming, etc.)

Thanks, Ben

bchnge commented 10 years ago

Hey ya'll,

Just touching base.

I think we are going to meet after class on Monday, given that this weekend is Thanksgiving. As of now, I think everyone knows what they need to do. In particular:

Kyle will have a police precinct level data set of variables such as number of forceful stops, successful stops, etc.

Arpit will continue refining the census-tract level k-means results. Would be useful to also provide some interpretation of the results and a 'dictionary' of what each run entails (which variables were used, etc.)

Davar will have a police precinct level data set of crime data that Ruxanda can use. (i did something quick in excel using the data found in the repo and i got a negative number in one or two of the precincts...is the data describing counts or relative change in number of crimes committed?)

I think it's important, aside from having these data tables ready by early next week, is to begin to have some general idea of what our results are. So, on Monday, in addition to having the data tables ready for merging, I'd like to chat about some interesting/non interesting things each of you have found in the pieces you've done. Do we see big skews in a particular precinct or segment type from the kmeans? If so, where. I think we will begin to have a lot of 'results' as the data now begin to come together, but we need to also begin to understand these results to better determine what to present in the final product. I think the best way to achieve this is for each person to take ownership (not just in data creation, but also in insight) of their own task. if anyone's unsure of what they need to do, now's the time to speak up.

As for myself, I've finally made some decent progress with D3 mapping. Check out test_pp.html (do a full pull). We have color gradients and tooltips based off an external csv working (the current examples uses some basic numbers from the crime data Davar has uploaded. My next steps are too incorporate some more interactiveness (zoom in/out, drop downs, etc.)...will let you know how things go.

Okay, time for sleep!

Cheers, Ben