Regression

source

Regression(
   data_obj: Regression_Data
)

Class for all regression methods

Methods:

.process_data

source

.process_data(
   data_obj
)

Drop non-numeric columns from the dataframe, tries to find a column with date and converts it to usable format

Args

  • data_obj : Regression_Data object

Returns

modified self.data and data_obj.result_string to inform user of processing

.split_data

source

.split_data(
   data_obj
)

Splits given dataset into evidence and labels, requires labels to be last column of dataframe

Args

  • data_obj : Regression_Data object

Returns

modified data_obj

source

.print_results(
   data_obj
)

Adds the results to the result_string for the GUI

Args

  • data_obj : Regression_Data object

Returns

modified data_obj

.plot_predictions

source

.plot_predictions(
   y_scaler, y_test, predictions, data_obj, train_test
)

Plots the predicted and real values against each other. Plots as many values as the user selected

Args

  • y_scaler : MinMaxScaler used to scale y-data
  • y_test : Real values for y
  • predictions : Predicted values for y
  • data_obj : Regression_Data object
  • train_test : str "train" or "test" for plot title and correct output to data_obj

Returns

modified data_obj