This repository contains visualizations I use for analysis / some papers. These are the abstract functions that should be generally applicable to any data frames. For more in-depth explanations please refer to the respective functions.
If you have make installed, setup the virtual environment using:
make venv
There are some example scripts that plot random / fantasy data under examples/. Before
you run any of these please activate the virtual environment using:
source vispp_venv/bin/activate
The plot_matched function can be used to create
something similar to a strips / scatter plot to compare the performance of, e.g.,
different classifiers on the same dataset, as long as the number off data points is
somewhat limited.
The added benefit of this plot function is that you can compare / track a single subject across the classifiers. This is achieved by selecting same markers for the same data point and a consistent x-position within one strip. Usually you want are interested mostly in the performance in one classifier (i.e., your fancy new method). Therefore you can sort the x-position according to the performance of your fancy new method. For example:
In this plot, within-strip position is sorted by the classification score obtained by the Fancy method. You can somewhat easily track average performance, standard deviation as well as the classification score for each individual subject across methods.