search

CCS-Lab / easyml

A toolkit for easily building and evaluating machine learning models.
https://ccs-lab.github.io/easyml
Other
40 stars 16 forks source link

alpha in easy_glmnet as an input #56

Closed youngahn closed 7 years ago

youngahn commented 7 years ago

alpha (between 0 (ridge) and 1 (lasso)) should be an input.

See ?glmnet::glmnet and 'alpha' (default = 1).

paulhendricks commented 7 years ago

The current implementation should allow for any arguments to glmnet (or random forest or support vector machines for that matter) to be passed through, in both Python and R.

For example, the below code should work in R without any problem:

library(easyml) # https://github.com/CCS-Lab/easyml

# Load data
data("cocaine_dependence", package = "easyml")

# Analyze data
results <- easy_glmnet(cocaine_dependence, "diagnosis", 
                       family = "binomial", exclude_variables = c("subject"), 
                       categorical_variables = c("male"), preprocessor = preprocess_scaler, 
                       n_samples = 100, n_divisions = 100, n_iterations = 10, 
                       random_state = 1, n_core = 8, alpha = 1)

Was there a specific behavior you noticed that didn't work properly or was this just a generic request?

youngahn commented 7 years ago

Okay great. It wasn't clear any arguments to glmnet can be passed through. Later please make a note in Description.

youngahn commented 7 years ago

But I think it's a good idea to leave 'alpha' as an input argument in easy_glmnet. It's an important tuning parameter of glmnet and users need to know its existence.