ilanman

This is my first foray in d3. It's a force directed graph representing correlations amongst college students' vices. Thicker bands represent stronger correlations, and larger nodes represent larger proportions of student partaking in that vice. More info found on my blog.

ilanman

This chart compares congressional workload over the years to the appropriations they are given. There is also the option of looking at the U.S. CPI over the same time to highlight the discrepancy. I am using d3-tip for my tooltips. More context around this chart can be found on my blog post.

ilanman

This is a simple visualization of customer landings to a website due to advertisement. Project forward the time series prior to advertisement, then calculate the difference between the actual landings to the projected. See blog for more details.

Looking for suggestions to improve code and/or visual design.

ilanman

This graph looks at how the Senate and House Democrats and Republicans have voted with or against the President over the years. It's clear what the President's party is by Congress' voting patterns. More info here. Main inspiration was this.

ilanman

This is an update to a ggplot version of the same graphic. It's looking at more popular social media mentions in various NY Times sections, over time. More info here.
Main inspiration was this

ilanman

This is a heatmap inspired by this Trulia recreation, with the following additions:

1. This iteration uses [colorBrewer.js](https://github.com/mbostock/d3/blob/master/lib/colorbrewer/colorbrewer.js) for the color scheme.
2. Tool-tips using [tool-tips](http://labratrevenge.com/d3-tip/) to show the value.
The user can update the tile values. The values are generated from a probability distribution. I created a short script (rand.js) that generates a Normal random variable using the Box-Muller transformation, an Exponential distribution and a Logistic distribution.

ilanman

The US government released medicare data in early 2013. This is an analysis of that data across the US. It compares the cost of treatments per hospital and state vs. total medicare coverage and shows where there is greatest discrepancy.

ilanman

This visualization generates paths of probability distributions from the heatmap example. It also uses Twitter Bootstrap for styling.

Next iteration will include input from user to change mu (mean), sigma (standard deviation), lambda (rate parameter for an exponential, x_m and alpha for the pareto and alpha and beta for the gamma.

ilanman

This is a choropleth of Medicare payments by state, using the data recently released by the Center for Medicare and Medicaid Services. I used a kmeans algorithm to determine that 3 clusters explains roughly 90% of the variation in the data.

ilanman

This gist has the code to visually display the results of a binary classifier on tennis statistics, using an interactive ROC curve.