Example data preparation ASCAT - ISMN

The validation framework allows users to prepare their data before the validation is performed by using the DataPreparation class which must contain two methods:

prep_reference, has at least one parameter - the reference dataframe
prep_other, has at least two parameters - the other dataframe and the name of the dataset



In [ ]:

    
class DataPreparation(object):
    """
    Class for preparing the data before validation.
    """
    @staticmethod
    def prep_reference(reference):
        """
        Static method used to prepare the reference dataset (ISMN).

        Parameters
        ----------
        reference : pandas.DataFrame
            ISMN data.

        Returns
        -------
        reference : pandas.DataFrame
            Masked reference.
        """
        return reference

    @staticmethod
    def prep_other(other, other_name, mask_snow=80, mask_frozen=80, mask_ssf=[0, 1]):
        """
        Static method used to prepare the other datasets (ASCAT).

        Parameters
        ----------
        other : pandas.DataFrame
            Containing at least the fields: sm, frozen_prob, snow_prob, ssf.
        other_name : string
            ASCAT.
        mask_snow : int, optional
            If set, all the observations with snow probability > mask_snow
            are removed from the result. Default: 80.
        mask_frozen : int, optional
            If set, all the observations with frozen probability > mask_frozen
            are removed from the result. Default: 80.
        mask_ssf : list, optional
            If set, all the observations with ssf != mask_ssf are removed from
            the result. Default: [0, 1].

        Returns
        -------
        reference : pandas.DataFrame
            Masked reference.
        """
        if other_name == 'ASCAT':

            # mask frozen
            if mask_frozen is not None:
                other = other[other['frozen_prob'] < mask_frozen]

            # mask snow
            if mask_snow is not None:
                other = other[other['snow_prob'] < mask_snow]

            # mask ssf
            if mask_ssf is not None:
                other = other[(other['ssf'] == mask_ssf[0]) |
                              (other['ssf'] == mask_ssf[1])]
        return other