Composite Regression

multi_regression.DTRTAV(file_contents, data_type, regression_var)

‘Do The Right Thing’ using adapative-validated glmnet as the regression method.

Pick the right version of glmnet to use based on data type.

Args:

file_contents: Raw form-data from HTTP POST request. data_type: Parameter of the same name from POST request. regression_var: Parameter ‘regression_index’ from POST request.

Returns:

Results from running glmnet on input data, this will be a pandas DataFrame containing the result cofficients, including labels.

Raises:

ValueError: Requested data_type is not currently supported, valid types are ‘csv’ ‘xlsx’ and ‘json’.

multi_regression.DTRTCV(file_contents, data_type, regression_var)

‘Do The Right Thing’ using cross-validated glmnet as the regression method.

Pick the right version of glmnet to use based on data type.

Args:

file_contents: Raw form-data from HTTP POST request. data_type: Parameter of the same name from POST request. regression_var: Parameter ‘regression_index’ from POST request.

Returns:

Results from running glmnet on input data, this will be a pandas DataFrame containing the result cofficients, including labels.

Raises:

ValueError: Requested data_type is not currently supported, valid types are ‘csv’ ‘xlsx’ and ‘json’.

multi_regression.DTRTFOS(file_contents, data_type, regression_var)

‘Do The Right Thing’ using FOS as the regression method.

Pick the right version of FOS to use based on data type.

Args:

file_contents: Raw form-data from HTTP POST request. data_type: Parameter of the same name from POST request. regression_var: Parameter ‘regression_index’ from POST request.

Returns:

Results from running FOS on input data, this will be a pandas DataFrame containing the result cofficients, including labels.

Raises:

ValueError: Requested data_type is not currently supported, valid types are ‘csv’ ‘xlsx’ and ‘json’.

multi_regression.MultiRegression(file_contents, regression_type, data_type, regression_var)

All-in-one regression function that allows regression method to be switched.

Args:

file_contents: Raw form-data from HTTP POST request. regression_type: The type of regression method that should be used ( eg FOS ) data_type: Parameter of the same name from POST request. regression_var: Parameter ‘regression_index’ from POST request.

Returns:

Results from running glmnet on input data, this will be a pandas DataFrame containing the result cofficients, including labels.

Raises:

ValueError: Requested regression_type is not currently supported. Valid method are ‘fos’ ‘cv’ and ‘av’ corresponding to FOS, glmnet w/ cross validation validation, and glmnet w/ adaptive validation respectively.