Skip to content

Data Scientist Associate Practical Exam Submission

GoalZone Booking Attendance Analysis

Task 1

The dataset contains 1500 rows and 8 columns with some missing values before cleaning. I have validated all the variables against the set criteria in task description:

  • booking_id: same as described
  • month_as_member: same as decribed and no missing values
  • weight: same as decribed and with 20 missing values and filled the missing values with the average of the column
  • days_before: no missing values and used tring replace method to convert to remove the days affix in some of the values
  • day_of_week: same as decribed and no missing values
  • time: same as decribed and no missing values
  • category: same as decribed and no missing values

After the data validation, the dataset contains the same number of records and fields

Loading original Dataset

# Data Validation

# import the reuired libraries

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
%matplotlib inline

# data validation - all variables are as decribed in the task list
df = pd.read_csv('fitness_class_2212.csv')
df.info()
# check the number of missing values
df.isna().sum()

Validate the categorical variables

cat_columns = ['day_of_week', 'time', 'category']
# validate the minimum values of the columns months_as_member, days_before, weight
print("Minimum values")
for col in ['months_as_member', 'days_before', 'weight']:
    print(f"{col}: {df[col].min()} ")

Filling the missing values

# fill missing values with predefined values
def fill_missing(df):
    to_be_unknown, to_be_avg, to_be_0 = ['day_of_week', 'time', 'category'], ['months_as_member', 'weight'], ['days_before']
    for col in df.columns:
        if col in to_be_unknown:
            df[col].fillna('unknown', inplace=True)
        if col in to_be_avg:
            df[col].fillna(df[col].mean(), inplace=True)
        if col in to_be_0:
            df[col].fillna(0, inplace=True)
    df.dropna(inplace=True) # if not in either ie column attended, remove the row
    return df
df = fill_missing(df)
# check values if they fit description
for col in ['day_of_week', 'time', 'category']:
    print(f"{col} \n {df[col].unique()}")
df['days_before'].unique()

Cleaning the data

# clean values to fit description - day_of_week(wed, mon, fri), category(-, to unknown)
def clean_day_of_week(df):
    df['day_of_week'].mask(df['day_of_week'].str.startswith('Wed'), 'Wed', inplace=True)
    df['day_of_week'].mask(df['day_of_week'].str.startswith('Mon'), 'Mon', inplace=True)
    df['day_of_week'].mask(df['day_of_week'].str.startswith('Fri'), 'Fri', inplace=True)
    return df

def clean_category(df):
    df['category'].mask(df['category'] == '-', 'unknown', inplace=True)
    return df

# clean types - from object to pandas category for easier working with during visualisation and modelling
def clean_types(df):
    for col in cat_columns:
        df[col] = df[col].astype('category')
    df['days_before'] = df['days_before'].astype('int64')
    return df

# claen days_defore
def clean_days_defore(df):
    df['days_before'] = df['days_before'].str.replace(' days', '')
    return df

df = clean_day_of_week(df)
df = clean_category(df)
df = clean_days_defore(df)
df = clean_types(df)
df.info()
‌
‌
‌