Skip to content
Deshan Perera edited this page Nov 21, 2024 · 20 revisions

In this section we will look at how to install CATE

C/C++ CUDA CI

CATE can be used on-device (installation time usually less than 5 minutes) or online using Google Colab.

CATE on Google Colab

Open In Colab

To use CATE online via Google Colab please use the link: CATE on Colab.

Please note to use a Colab session with a CUDA-enabled GPU (E.g. T4 GPU).

On device installation via Anaconda

Anaconda-Server Badge Anaconda-Server Badge Anaconda-Server Badge

To install CATE using the Anaconda package execute the following command.

conda install deshan_cate::cate

Test run

Try running the executable by typing the following (this will print the help menu):
CATE -h

On device installation via GitHub

To install CATE as is you will have to download the code from our GitHub repository and compile it on-site.

The recommended command line prompts are provided below.

Download the repository

Download the repository using the following command:
git clone "https://github.com/theLongLab/CATE/"

Navigate into the downloaded folder:
cd CATE/

Load the modules

In some instances you will have to have both the GCC and NVCC compilers loaded, usually having the nvcc compiler loaded is sufficient.

Load the GCC compiler:
module load gcc/10.2.0

Load the NVCC compiler:
module load cuda/11.3.0

Compile the code

Compile the code using:
nvcc -std=c++17 *.cu *.cpp -o "CATE"

Test run

Try running the executable by typing the following (this will print the help menu):
./CATE -h