NeRd

In my last post I’ve introduced an implementation of a Substitution Rule Mining algorithm. Now I’ve developed a new version for better performance. Thus I used the purrr and tibble package from the tidyverse. Tibbles need much less memory like normal data frames and the purrr package contains fast... [Read More]

Imagine being a seller who wants to know what his customer often buy together. He wants some rules like “When people buy X they also often buy Y”, to get to know how e.g. he has to build up his store, so customers can find all they want very easily.... [Read More]

Today I’ve learned a little trick that probably will save much of my time and nerves. Imagine you would like to transform your data or select some of it and you could do it very easily with a for loop. But you need every of this transformation in a new... [Read More]

Some time ago I was frustrated by analyzing my first selfmade data set on the university, because of the dance of the p-values. But not only p-values were frustrating me but also the task from superiors to test a Nullhypothesis, which I hade to realize was not possible by... [Read More]

In most psychological questionnaires are reversed items, usually to test if people filled it correctly and read the items. But when we want to analyze the results of the questionnaire we need to be sure, that this negative items are reversed. Let’s look at an example. I’ve created four items... [Read More]