Recently, I had the opportunity to showcase tidymodels in workshops and talks. Because of my vantage point as a user, I figured it would be valuable to share what I have learned so far. Let’s begin by framing where tidymodels fits in our analysis projects.
The diagram above is based on the R for Data Science book, by Wickham and Grolemund. The version in this article illustrates what step each package covers.
We were recently presented with a problem where the decision maker wanted to understand how their data would naturally group together. The classic technique of k-means clustering was a natural choice; it’s well known, computationally efficient, and implemented in base R via the kmeans() function.
Our problem has a slight wrinkle: the decision maker wished to see the data grouped with (nearly) equal sizes. Now, a ‘true’ statistician would tell the client that the right thing to do from a theoretical perspective was to use native k-means results because some centers can simply have more nearby points than other centers.
Roland Stevenson is a data scientist and consultant who may be reached on Linkedin.
reticulate is an R package that allows us to use Python modules from within RStudio. I recently found this functionality useful while trying to compare the results of different uplift models. Though I did have R’s uplift package producing Qini charts and metrics, I also wanted to see how things looked with Wayfair’s promising pylift package.
Research design consists of a set of choices about what research procedures to use. For example, how many subjects to interview, which questions to ask them, and what to do in the analysis phase with the data that results from these choices. We do not have good tools for assessing whether the chosen procedures are good ones. DeclareDesign is an R package for learning about, implementing, and communicating research procedures, from data collection to data analysis.
One hundred eighty-seven new packages made it to CRAN in April. Here are my picks for the “Top 40”, organized into ten categories: Biotechnology, Data, Econometrics, Machine Learning, Medicine, Science, Statistics, Time Series, Utilities, and Visualization.
Biotechnology genpwr v1.00: Provides functions for power and sample size calculations for genetic association studies allowing for mis-specification of the model of genetic susceptibility. The methods employed are extensions of Gauderman (2002) and Gauderman (2002).
After an extended hiatus, Reproducible Finance is back! We’ll celebrate by changing focus a bit and coding up an investment strategy called Momentum. Before we even tiptoe in that direction, please note that this is not intended as investment advice and it’s not intended to be a script that can be implemented for trading.
This is part 4 of a four-part series about the HIV epidemic in Africa. In this final part, we discuss how genetic diversity can be used to classify laboratory samples into either inter-patient or intra-classes, using logistic regression. This helps with quality in the lab, since it’s possible to match new samples with samples from the same patient, taken years apart and allowing for mutation of the HIV virus genomic sequence.
This is part 3 of a four-part series about the HIV epidemic in Africa. In a recent publication in PLoS ONE, the authors described how they used affordable hardware to create a phylogenetic pipeline, tailored for the HIV drug-resistance testing facility. In this part, we discuss genetic diversity and how this can be analysed using markov chains and heatmaps.
Enjoy a virtual mushroom hunt with R and RSelenium, which allows R to use a web browser as a human would, including clicking on buttons, etc.
This is part 2 of a four-part series about the HIV epidemic in Africa. In a recent publication in PLoS ONE, the authors described how they used affordable hardware to create a phylogenetic pipeline, tailored for the HIV drug-resistance testing facility. Part 2 discusses drug-resistance testing of HIV isolates in sub-Saharan Africa.