Globally local variable importance
Project description
- Two measures to calculate local variable importance with global models. It means that calcuate variable importance for subset of global data set using global model trained by global data set. I provide two files including 'todi.py' and 'meda.py' to realise these two methods. Both file include a class called 'lovim' inheriting from 'RandomForestRegressor' in 'scikit-learn'.
- For example:
- l = lovim(500, max_features=0.3, n_jobs=-1))# just be same with the class 'RandomForestRegressor' in 'scikit-learn'.
- l.fit(X,Y)# fitting the model.
- local_variable importance = l.compute_feature_importance(X,Y,partition_feature = partition_feature, norm=True,n_jobs=-1) # calculate local variable importance.
- Of course, we upload the file to 'https://pypi.org/'. You can install it just like other packages such as numpy and pandas. Please input 'pip install glvi' in "cmd" for windows or Terminal Emulator for Linux to install the package.
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
glvi-0.1.4.tar.gz
(6.4 kB
view hashes)