pylusat.rescale module

pylusat.rescale.reclassify(input_df, input_col, reclassify_def, output_col, nodata=None)

Reclassify values in an existing column based on key-value pairs provided in a dictionary.

The function can handle both categorical and interval definitions. For interval definition, the keys of the dictionary should be a tuple of two numbers corresponding to the start and end of each interval. The intervals are right closed.

Parameters:

  • input_df (pandas.DataFrame or geopandas.GeoDataFrame) – Input DataFrame with the column need to be reclassified.

  • input_col (str) – The name of the input column containing the old values.

  • reclassify_def (dict) – The dictionary consists of definitions to convert old values to new values.

  • output_col (str) – The name of the output column.

  • nodata (int or float, optional) – The value used to fill the nodata records.

Returns:

input_df – The output DataFrame with the reclassified values.

Return type:

DataFrame or geopandas.GeoDataFrame

pylusat.rescale.linear(input_df, input_col, output_col, start=None, end=None, output_min=1, output_max=9)

Rescale a column in a DataFrame linearly.

If argument start is greater than end, the rescaling is in the same direction as values in the input column, i.e., smaller (bigger) values in the input column correspond to smaller (bigger) values in the output. If argument start is less than end, the rescaling is in the reverse direction as values in the input column. The start and end of the input column do not necessarily to be the minimum and maximum of the input column. Values beyond the specified bound will be assigned to output_min and output_max, depending on which side they are on.

Parameters:

  • input_df (pandas.DataFrame or geopandas.GeoDataFrame) – Input DataFrame containing a column need to be rescaled.

  • input_col (str) – Name of the old column.

  • output_col (str) – Name of the new column.

  • start (int or float) – Value from which the rescaling starts.

  • end (int or float) – Value from which the rescaling ends.

  • output_min (int or float) – The minimum value of the output column.

  • output_max (int or float) – The maximum value of the output column.

Returns:

input_df – Output DataFrame containing the rescaled column.

Return type:

pandas.DataFrame or geopandas.GeoDataFrame

Examples

Linearlly rescale the enrollment column of the schools GeoDataFrame.

>>> pylusat.rescale.linear(schools_gdf, "ENROLLMENT", "ENROLL_CLS")
                                        NAME    ENROLL_CLS
0               COUNTRYSIDE CHRISTIAN SCHOOL    1.384930
1     TRILOGY SCHOOL OF LEARNING ALTERNATIVE    1.305039
2               MILLHOPPER MONTESSORI SCHOOL    1.784385
3              ST MICHAEL'S EPISCOPAL SCHOOL    1.000000
4                     BNAI ISRAEL DAY SCHOOL    1.079891
...
116     PERSIMMON EARLY LEARNING ACADEMY LLC    1.000000
117 BUSY BEE AND BUTTERFLY CHRISTIAN ACADEMY    1.000000
118                       THE PHENOM ACADEMY    1.000000
119     ALACHUA LEARNING CENTER, INC. MIDDLE    1.221516
120                FLORIDA SCHOOL OF MASSAGE    1.000000
pylusat.rescale.gamma(input_df, input_col, output_col, output_min=1, output_max=9)

Rescale values in a column based on an asymptotic gamma distribution.

The function calls a gamma object from the stats module in the SciPy package, see more detail from: https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.gamma.html First, Maximum Likelihood Estimation (MLE) is used to estimate the shape and scale parameters from the data. Then, the function calculates a new value for its associated old value, x, by scaling the probability of the random variable being less than or equal to x, based on the the estimated cumulative density function (cdf).

Parameters:

  • input_df (pandas.DataFrame or geopandas.GeoDataFrame) – Input DataFrame containing a column need to be rescaled.

  • input_col (str) – Name of the old column.

  • output_col (str) – Name of the new column.

  • output_min (int or float) – The minimum value of the output column.

  • output_max (int or float) – The maximum value of the output column.

Returns:

input_df – Output DataFrame containing the rescaled column.

Return type:

pandas.DataFrame or geopandas.GeoDataFrame