Installation

We recommend to install HTS in a virtual environment.

HTS uses GPy for Gaussian process normalization. GPy itself requires numpy at setup time:

$ pip install numpy

With pip (or pip3) configured for Python3, you can install the latest version of HTS directly from Github.:

$ pip install git+https://github.com/elkeschaper/hts.git

GPy is under constant development, and it might be useful to deinstall the PyPi version installed by default, and instead install the latest develop version:

$ pip install git+https://github.com/SheffieldML/GPy.git

Now you can import hts in your Python3 project:

import hts

For creating .html or .pdf reports with HTS, install Pandoc (v > 1.12.3) and an up-to-date version of R (v >= 3.2.4). Make sure that Rscript is accessible from the command line:

$ which Rscript

Next, within R, install these R packages:

install.packages("ggplot2");
install.packages("gridExtra");
install.packages("knitr");
install.packages("MASS);
install.packages("rmarkdown");

We use Knitr to create the reports from R snippets.