Skip to content

Data Analyst Certification Case Study

Office Supplies
  • Pens & Printers is a national office supplies chain. At the moment, they send office supplies out of warehouses in four regions: East, West, South, and Central, and all four warehouses stock the same products. The Head of Sales thinks this leads to large amounts of unsold products in some locations. The management at Pens & Printers would like to look at the data and present to them the popularity of products in each region. Are there products that do not sell in some locations? Are there any other patterns over time in each region that you can find in the data?

1. Data Overview & Wrangling

# Import needed libraries for analysis 
import numpy as np 
import pandas as pd 
import matplotlib.pyplot as plt 
import seaborn as sns 
%matplotlib inline 
# Load data from csv 
df = pd.read_csv('data/office_supplies.csv')
df.head()
# explore data & get overview 
df.info()
df.describe()
# modify data types of different columns 
df['Order Date'] = df['Order Date'].astype('datetime64[ns]')
for column in df.columns : 
    if df[column].dtype == 'O': 
        df[column] = df[column].astype('string')
df.info()
# update all data columns title to be lower case and without spaces 
df.columns = df.columns.str.lower()
df.columns = df.columns.str.replace(' ', '_')
df.columns = df.columns.str.replace('-', '_')
df.info()
# fill null values with zero 
df.fillna(0, inplace=True)
# add new column year 
df['year'] = df['order_date'].dt.year
df.info()

2. Exploratory Data Analysis

2.1 Univariate EDA
# view distribution of sales in different regions 
sns.color_palette()[0]
fig, axes = plt.subplots(nrows=1, ncols=4, figsize=(30, 10))
regions = df['region'].unique()
for ax, region in zip(axes, regions):
    sns.boxplot(data=df[df['region']==region], x='sales', ax=ax).set(title=f'{region}')
# view distribution of profits in different regions 
fig, axes = plt.subplots(nrows=1, ncols=4, figsize=(30, 10))
regions = df['region'].unique()
for ax, region in zip(axes, regions):
    sns.boxplot(data=df[df['region'] == region], x='profit', ax=ax).set(title=f'{region}')
# View count of sales transactions for each category in each region
fig, axes = plt.subplots(nrows=1, ncols=4, figsize=(30, 10))
regions = df['region'].unique()
for ax, region in zip(axes, regions):
    sns.countplot(data=df[df['region']==region], x='category', ax=ax).set_title(f'{region}')
# View count of sales transactions for each sub-category in each region
fig, axes = plt.subplots(nrows=1, ncols=4, figsize=(30, 10))
regions = df['region'].unique()
for ax, region in zip(axes, regions):
    sns.countplot(data=df[df['region']==region], x='sub_category', ax=ax).set(title=f'{region}')
    ax.tick_params(axis='x', rotation=60)
‌
‌
‌