preprocessing

• DataSet
  • DataSet
  • DataSetUnion
  • DataSetPipeline
• Imputation
  • Imputation
• Featurization
• HRU Discretization
  • HRU Discretization
• Data Transformations
  • Transformations

The preprocessing sub-module contains classes which handles preparation of input data. The fundamental class is the DataSet class which prepares data from a single data source. If you hvae multiple data sources then you can either use DataSetUnion or DataSetPipeline class. The DataSet can take a data in a variety of commonly found formats such as csv, xlsx and prepares the data so that it can be fed to Model for training. This class works with modules in conjunction with Imputation class.

It should be noted that transformations applied in ai4water are part of Model. This means transformations are applied, everytime a call to the model is made using fit, predict, evaluate, score or predict_prob methods.