microbiota_host_classifier

A random forest model to detect pet samples in human microbiome data

To repeat analysis

unpack data.tar.gz
make sure the following Python modules are installed: pandas==0.25.1 numpy==1.17.2 scikit-learn==0.21.3 scikit-bio==0.5.5 scipy==1.3.1 statsmodels==0.10.1 matplotlib==3.1.1 seaborn==0.9.0 joblib==0.13.2 fpdf==1.7.2
run ./launch.sh script in the directory with 'data' folder, it will make folders with results (tables and figures) and joblib objects (fitted RF models) (note that for for nbconvert to produce output .pdf files the TeX should be also installed)

Name		Name	Last commit message	Last commit date
Latest commit History 10 Commits
README.md		README.md
catsNdogs_part0_Load_data.ipynb		catsNdogs_part0_Load_data.ipynb
catsNdogs_part1_MDS.ipynb		catsNdogs_part1_MDS.ipynb
catsNdogs_part2_MW_test.ipynb		catsNdogs_part2_MW_test.ipynb
catsNdogs_part3_2_RF_feature_importance.ipynb		catsNdogs_part3_2_RF_feature_importance.ipynb
catsNdogs_part3_RF_v2.ipynb		catsNdogs_part3_RF_v2.ipynb
catsNdogs_part4_PJC.ipynb		catsNdogs_part4_PJC.ipynb
catsNdogs_part5_Test_new_projects.ipynb		catsNdogs_part5_Test_new_projects.ipynb
data.tar.gz		data.tar.gz
launch.sh		launch.sh
shared_functions.ipynb		shared_functions.ipynb

Provide feedback