Skip to content

Latest commit

 

History

History
 
 

1.split_data

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 
 
 
 
 
 
 

1. Split Data

In this module, we split the training data into training and testing datasets.

Data is split into subsets in split_data.ipynb. The testing dataset is determined by randomly sampling 15% (stratified by phenotypic class) of the single-cell dataset. The training dataset is the subset remaining after the testing samples are removed. We store sample indexes associated with training and testing subsets in indexes/, and we later use these sample indexes to load subsets from labeled data in 0.download_data/data/.

Step 1: Split Data

Use the commands below to create indexes for training and testing data subsets:

# Make sure you are located in 1.split_data
cd 1.split_data

# Activate phenotypic_profiling conda environment
conda activate phenotypic_profiling

# Split data
bash split_data.sh