Data.normalise

Data.normalise(target=None, base=None, replace=True, header=None, scale=None, limits=(0.0, 1.0))

Normalise data columns by dividing through by a base column value.

Parameters:

target (index) – One or more target columns to normalise can be a string, integer or list of strings or integers. If None then the default ‘y’ column is used.

Keyword Arguments:
  • base (index) – The column to normalise to, can be an integer or string. Deprecated can also be a tuple (low, high) being the output range

  • replace (bool) – Set True(default) to overwrite the target data columns

  • header (string or None) – The new column header - default is target name(norm)

  • scale (None or tuple of float,float) – Output range after normalising - low,high or None to map to -1,1

  • limits (float,float) – (low,high) - Take the input range from the high and low fraction of the input when sorted.

Returns:

(Stoner.Data) – The newly modified Data object.

Notes

The limits parameter is used to set the input scale being normalised from - if the data has a few outliers then this setting can be used to clip the input range before normalising. The parameters in the limit are the values at the low and high fractions of the cumulative distribution function of the data.