COVID research

import pandas as pd

import seaborn as sns

# select just one resuls_source file

# --> EDIT

# results_source = pd.read_csv(_, index_col="ID odpowiedzi")

# results_source = pd.read_csv(_, index_col="ID odpowiedzi")

# results_source = pd.read_csv(_, index_col="ID odpowiedzi")

results_source = pd.read_csv(
results_institutions.csv, index_col="ID odpowiedzi")

# results_source = pd.read_csv(_, index_col="ID odpowiedzi")

# don't edit

def filter_city(results, city='All'):

    if city in ('Katowice', 'Lublin', 'Wrocław'):

        mask = results.columns[results.columns.str.startswith('W jakim mieście')][0]

        results = results[results[mask]==city]

        return results

    elif city is 'All':

        return results

    else:

        raise ValueError('Value needs to be one of: "All / Katowice / Lublin / Wrocław"')

def aggr_in_cols(df, first_col, last_col, normalize):

    result = df.iloc[:, first_col:last_col + 1].apply(pd.Series.value_counts, normalize=normalize)

    if normalize:

        result = result.applymap(lambda x: round(x * 100, 2))

    return result

# CHECK COLUMN NUMBER

for idx,r in enumerate(results_source):

    print(idx,r)

# --> FILTER VALUES FOR ALL OR ONE CITY

city = 'All' # e.g. city = 'Katowice'

# don't edit

results = filter_city(results_source, city)

# --> SELECT VARIABLES TO COMPARE (CHECK THE COL NUM ABOVE)

col1 = 202

col2 = 0

# don't edit

crosst = pd.crosstab(results[results.columns[col1]],

            results[results.columns[col2]],

            margins=True,

            normalize='columns',  # or: 'columns'

           )

crosst = crosst.applymap(lambda x: round(x * 100, 2))

crosst

# Generate visualisation

viz = sns.heatmap(crosst, cmap="YlGnBu", annot=True, cbar=False)

viz.set_ylabel('col1')

viz.set_xlabel('col2')

viz.plot()

# % for values in multiple columns

# insert col numbers here

start = 1

end = 5

percent = True

aggr_in_cols(results, start, end, percent)