kailo_beewell_dashboard.synthesise_scores

Functions analysing scores for each topic - as part of several files which provide functions for synthesis (creation and aggregation) of data for the dashboard.

Module Contents

Functions

sum_score(df)	Find the sum of the provided columns. If any of the required columns
reverse_score(scores, min, max)	Reverse scores in the provided array, based on the known min and max of the
calculate_scores(data)	Creates scores for each pupil in the provided dataframe, for each of the
score_descriptives(values, counts)	This function uses site-level data. Using the mean and count from each
create_rag_ratings(df)	Generate rag ratings (above, average, below) based on scores

kailo_beewell_dashboard.synthesise_scores.sum_score(df)

Find the sum of the provided columns. If any of the required columns contain, NaN, it will just return NaN as the result

Parameters

dfpandas DataFrame

Dataframe just containing the columns you want to sum

kailo_beewell_dashboard.synthesise_scores.reverse_score(scores, min, max)

Reverse scores in the provided array, based on the known min and max of the scale of the scores. NaN will remain as NaN.

Parameters

scoresarray

Array with scores to be reversed

minint

Minimum possible score

maxint

Maximum possible score

Returns

Array with scores reversed

kailo_beewell_dashboard.synthesise_scores.calculate_scores(data)

Creates scores for each pupil in the provided dataframe, for each of the survey topics. Note, when referring to where scores are “set to positive” or “in a positive direction” or a “negative directioN”, this refers to whether the maximum score is a positive or negative outcome.

Parameters

datapandas dataframe

Pupil-level survey responses

Returns

datapandas dataframe

Pupil-level survey responses with the addition of topic scores

kailo_beewell_dashboard.synthesise_scores.score_descriptives(values, counts)

This function uses site-level data. Using the mean and count from each group, it calculates the weighted mean and weighted standard deviation of scores across the groups. It returns this, alongside a count of the pupils and groups included.

Additional information about weighted standard deviation: This normalises weights so they sum 1 (and so they can’t all be 0). It returns the biased variance and is like a weighted version of np.std(). For small samples, may want to alter to unbiased variance. Based on: https://stackoverflow.com/questions/2413522/weighted-standard- deviation-in-numpy

Parameters

valuespandas series

Dataframe column with the mean scores in each group, NaN removed

counts: pandas series

Dataframe column with the count of pupils in each group, NaN removed

Returns

resultpandas Series

Series with each of the calculations, where index if the name of the calculation

kailo_beewell_dashboard.synthesise_scores.create_rag_ratings(df)

Generate rag ratings (above, average, below) based on scores

Parameters

dfdataframe

Contains scores by site, and potentially by pupil group too

Result:

ragdataframe

Dataframe with scores by site, with additional columns providing RAG ratings and descriptives of the score distribution that were used to generate the RAG