Convert String Parameter to Pandas Object

Question

below I have a function where I would like the parameter "label" be transformed as dataframe column/label object. In the function, the parameter "label" is the string "alcohol". Once accessed in the function via the parameter label which is a string "alcohol", I need it be used as the name of the dataframe column/label. The dataframe column is named "alcohol" as well. A call such as df.label.median() should be equivalent to df.alcohol.median() where alcohol is an actual column in the dataframe.

import pandas as pd
df = pd.read_csv('winequality-red.csv', sep=';')

def mean_quality_rating(df, label):
    median_label = df.label.median() #should evaluate as df.alcohol.median()
    for i, the_label in enumerate(df.label):
        if the_label >= median_label:
            df.loc[i, label] = 'high'
        else:
            df.loc[i, label] = 'low'
    return df.groupby(label).quality.mean()

mean_quality_rating(df, 'alcohol')

Answer 1

Try:

def mean_quality_rating(df, label):
    median_label = df[label]median() #should evaluate as df.alcohol.median()
    for i, the_label in enumerate(df[label]):
        if the_label >= median_label:
            df.loc[i, label] = 'high'
        else:
            df.loc[i, label] = 'low'
    return df.groupby(label).quality.mean()

mean_quality_rating(df, 'alcohol')

You can't not use the dot notation with variables.