Repository to the final project of Computational Reproducible Research course from Unicamp 1st semester master course.
The files are organized as follow:
- /data - Dataset used, if not here download and put on the folder the train.csv file in https://zenodo.org/record/3924900#.XvwnkfJKg5k
- /deliver - The executable paper and experiment's codes
- /dev - Codes and notebooks used during the experiment
- /figures - Figures used on research
- /utils - Some recommendations about the process of research in a Best Practices file
To run our "Reproducible_Paper.ipynb" in "/deliver" folder you must first install the pre requisites:
You can run the file in ubuntu by installing the dependencies in "requeriments.txt"
So make sure you have python3 and pip3 installed
Git clone this repository
Go to the folder in terminal
Type
python3 -m pip install -r requeriments.txt
Then you should be able to open jupyter and run the notebook /deliver/Reproducible_paper.ipynb.
Install Anaconda with python 3 and make sure you have pip installed
Then clone the repository
Open the Anaconda Prompt
Go to the repository folder, where there is the "environment.yml"
Type "conda env create -f environment.yml" and then "conda activate reproducibility"
Done. You should be able to run the reproducible paper.
To run the entire pipeline you must run all cells of the following notebooks inside the conda env:
/deliver/Reproducible_Paper.ipynb
If you dont have GPU you can run in colab with the notebook: https://drive.google.com/file/d/1CSTXdHgFwJltWDmODIqx-HD7SqyTzjG8/view?usp=sharing
- Data: CC BY-NC-SA 3.0
- Code: GNU GPL 3.0