In this blog post I tackle a problem for which I have been looking for an of-the-shelve solution: Converting the columns of a certain type of class to another class, while preserving the data frame. To this day I haven’t found it. That’s why I wrote this function. Let’s say…
The following error message is something that me and my colleagues keep running into when using Rstudio. ‘Error Saving File — The device does not recognize the command’ There are multiple things that cause this, but basically it’s because RStudio is having troubles writing to the file. Here’s what you…
In this blog post, I elaborate on setting axis limits in a plot, generated by ggplot2. There are two ways: one where you pretend the data outside the limits doesn’t exist (using lims), and one where you respect that the data outside the limits exists (using coord_cartesian). The documentation for…
Let’s say you want to use principal component analysis on the numerical columns in your data set to reduce the amount of features in your model and get rid of multicollinearity. For that, you’d need to select the numerical columns only. Now how could you do that properly? In the…
Recently I bumped into an issue where a query, ran using the DBI library with the dbGetQuery function returned an integer64. Even without me noticing it. When converting my data frame (or data.table) to a matrix for clustering purposes, I ran into the following error: Error in if (changes ==…
K-means clustering is quick and dirty and generally provides some interesting results. However, the default kmeans function in R lacks features, such as actually storing the model to use the centroids for prediction purposes on unseen data. That’s where flexclust comes in. Flexclust is a package that is designed around…
One of the things that blew my mind the first time I learned about it is bootstrapping. Although very counterintuitive at first glance, it makes perfect sense to estimate the accuracy of your sample estimates, given that you have a representative sample. In this blog post I explain how you…
What’s really cool is that Keras in R now has an interface to AI Platform (fka Cloud ML) in Google Cloud Platform. You do some trial-and-error on your computer and once you’re good to go, you can submit your job with a couple lines of code. Just to be clear:…
If you too like Keras and RStudio, you’re probably curious how you can hypertune a model. Here goes. If you’ve paid attention in the previous part, you noticed that I didn’t do any hypertuning at all to tweak the performance of the car price prediction model. I tackle these questions…
In this blog post I will introduce you to building and training your own neural network algorithm in R through Keras & TensorFlow. If you haven’t installed Keras for R yet, please follow the instructions explained in part 1. I have explicitly chosen to work with structured data in this…