R Weekly 2017 Issue 4

23 Jan 2017

Highlight

• Statistical Thinking; a blog by Frank Harrell

• More components of the R ecosystem

• Introduction to Event Log Mining with R

Tutorials

• Learn Support Vector Machine (SVM) from Scratch in R

• weakly informative reparameterisations for location-scale mixtures

• R to the Rescue

• Structural Equation Modelling in R (Part 2)

• Reproducible Finance with R: Sector Correlations

• Regression model with auto correlated errors – Part 3, some astrology

• Git Gud with Git and R

• Two meanings of priors, part II: Quantifying uncertainty about model parameters

• R For Beginners: Basic Graphics Code to Produce Informative Graphs, Part Two, Working With Big Data

• Creating beautiful charts in chinese with ggplot2

• Feature Selection in Machine Learning (Breast Cancer Datasets)

• How to really do an analysis in R (part 1, data manipulation)

• Two meanings of priors, part I: The plausibility of models

• The Meeting Point Locator

• Gelman’s MrP in R – What is this all about?

• Repurposing the Candlestick Chart

• Beginners Exercise: Bayesian Computation with Stan and Farmer Jöns

• Regression model with auto correlated errors – Part 2, the models

• Rstudio Conference: Tips And Tricks

• R and Adobe Analytics: Two Dimensions, Many Metrics

Insights

• More components of the R ecosystem

• Data science opinions and tools to support them at rstudio::conf

• The Wizarding World of rstudio::conf 2017

R in the Real World

• Nobel Laureates – A Closer Look At The Data

• Workout Wednesday Redux (2017 Week 3) - This week’s workout is a challenge to reproduce or improve upon a chart by Matt Stiles.

• Diversity in the R Community

• GitHub Growth Appears Scale Free

• The Devil’s in the [Davos] Details — A quick look at this year’s WEF Global Risks Report

R in Organization

• Elements of a successful #openscience #rstats workshop

• What Benefits Are Companies Offering Developers?

Videos and Podcasts

• NSSD 31: The Word You’re Looking for is Magical - Hilary and Roger go on a podcast bender and discuss how they record the podcast, missing data methods, rstudio::conf, Uber Movement, diagnosis with prediction models, and Alexa/Siri.

• Introduction to Event Log Mining with R

Resources

• Data Import Cheat Sheet

• Data Transformation Cheat Sheet

• Statistical Thinking; a blog by Frank Harrell

New Packages and Tools

• prepdat - Preparing Experimental Data for Statistical Analysis

• ggedit - a GUI for advanced editing of ggplot2 objects

• VisualResume - Visual Resumes with R

• The Genetic Map Comparator - A user-friendly application to display and compare genetic maps

• padr - Quickly Get Datetime Data Ready for Analysis

• taskviewr - R Packages and Licenses by Task View

• CatterPlots - Did you ever wish you could make scatter plots with cat shaped points? Now you can!

• karyoploteR - Plot data along the genome ideograms

New Releases

• Rcpp 0.12.9 - Seamless R and C++ Integration

• Choroplethr 3.5.3 - choroplethr simplifies the creation of choropleths (thematic maps) in R

• RProtoBuf 0.4.8 - Windows support for proto3

• WEC 0.4 - an R package to apply ‘weighted effect coding’ to your dummy variables.

• rtweet 0.4 - Collecting Twitter data

Jobs

• R Jobs for R users – 12 jobs from around the world (2017-01-17)

R Project Updates

Updates from R Core:

• bw.SJ() and density(*, bw="SJ") no longer fail because of integer overflow for large sample size n (starting from n as small as 65536).

• download.file(method="libcurl") does not check for URL existence before attempting downloads; this is more robust to servers that do not support HEAD or range-based retrieval, but may create empty or incomplete files for aborted download requests.

• Bandwidth selectors bw.ucv() and bw.SJ() gave incorrect answers or incorrectly reported an error (because of integer overflow) for inputs longer than 46341. Similarly for b2.bcv() at length 5793. Another possible integer overflow is checked and may result in a reported error (rather than an incorrect result) for much longer inputs (millions for a smooth distribution).

Call for Participation

• R / Finance 2017 Call for Papers

• Call to join Forwards - the R Foundation taskforce on women and other under-represented groups.

• EARL 2017: Abstract submissions open for San Francisco and London + San Francisco dates and venue announced

Upcoming Events

• SatRday #2 Conference - Cape Town, South Africa 18 Feb, 2017
  The R community and some of South Africa’s most forward thinking companies have come together to bring satRday to Cape Town.

• useR! 2017 July 4, 2017
  The annual useR! conference is the main meeting of the international R user and developer community.

Quotes of the Week

p-value:chance someone else's data are more extreme than mine if H0 true, not the chance H0 is true given MY data https://t.co/HhfF9fwKLf

— Frank Harrell (@f2harrell) January 20, 2017

cleaning data? why not write your script for knitr::spin() so your coauthors can follow your journey #rstats pic.twitter.com/sH6NP74HxG

— Andrew MacDonald (@polesasunder) January 19, 2017