Publications by data_steve
d3/R Streamgraph on White House Petition Data
After last week’s post with the d3-htmlwidget chordplot showing the network of petition categories, in this post I decided to show the time trend aspect of how the petitions came in. I’m using @hrbrmstr’s htmlwidgets wrapper of a d3 streamgraph to help visualize the macro- and tag-specific trends. Compared to the chordplot, the time trend a...
2773 sym R (2576 sym/6 pcs) 2 img
d3/R Streamgraph on White House Petition Data
After last week’s post with the d3-htmlwidget chordplot showing the network of petition categories, in this post I decided to show the time trend aspect of how the petitions came in. I’m using @hrbrmstr’s htmlwidgets wrapper of a d3 streamgraph to help visualize the macro- and tag-specific trends. Compared to the chordplot, the time trend a...
2773 sym R (2575 sym/6 pcs) 2 img
Base R Nostalgia — by, tapply, ave, …
photo credit: Paul Yoakum This evening I was feeling nostalgic for base R group-bys. Before there was dplyr, there was apply and its cousins. I thought it’d be nice to get out the ol’ photo-album. To start off, the base R proto-ancestor of magrittr piping for me was the with function, especially with apply. It just cleaned up the syntax and v...
3104 sym R (1475 sym/10 pcs) 2 img
googleformr updates on CRAN
Main updates to googleformr have been to remove an internal dependency on magrittr chains in the function construction and to take http status checking outside of the main gformr function and into a check_form_works function. Both changes provide notable speed gains. One can still create a function with gformr as before: pacman::p_load(googlefor...
1722 sym R (148 sym/2 pcs) 2 img
Links I liked
The interwebs has been aflush with interesting data science material. Tweet Big Five Personality Tests Have Never Been So Fun(ny) Tweet Gelman on A Beautiful Sentiment Any time Gelman has something to say about graphics and methods, it’s useful. This seems like one of my first memories, though, of him discussing text analysis. T...
1273 sym 2 img
Need user feedback? Send it directly from R
Help Me, Help You As an R-dev, I’d like to know I’m meeting community need, as well as my own. But after an initial “aha!” moment, you don’t always know how to live up to the full potential of what you’ve started with a package. For example, RStudio collects extensive data through their products to better understand how to build out t...
1879 sym R (277 sym/1 pcs) 2 img
Links I liked
The interwebs has been aflush with interesting data science material. Tweet Big Five Personality Tests Have Never Been So Fun(ny) Tweet Gelman on A Beautiful Sentiment Any time Gelman has something to say about graphics and methods, it’s useful. This seems like one of my first memories, though, of him discussing text analysis. T...
1273 sym 2 img
feedbackr, R-devs + useRs
from Wikimedia In my last post I explained why I added a feedback function to googleformr. Since googleformr is an API for collecting data programmatically, I figured I’d leverage it for package development and accessibility. Below I’ll show how you can add your own feedback function to your package. Create your own feedback function First, m...
2234 sym R (452 sym/4 pcs) 2 img
feedbackr ⤌ R-devs + useRs
from Wikimedia In my last post I explained why I added a feedback function to googleformr. Since googleformr is an API for collecting data programmatically, I figured I’d leverage it for package development and accessibility. Below I’ll show how you can add your own feedback function to your package. Create your own feedback function First, m...
2234 sym R (452 sym/4 pcs) 2 img
Dumbbells get a little gg-smarter?
Bob Rubis @hrbrmstr recently has blogged several interesting posts [ 1, 2] about making dumbbell plots with ggplot2 via his package ggalt. @Tylerrinker did some really nice work reproducing a Pew dumbbell plot back in early 2015, which I found really helpful. Seeing @hrbrmstr’s posts reminded me of a project I did about 8 months ago. I was moc...
3467 sym R (2319 sym/3 pcs) 2 img