Why transformation reshapes data for analysis in Data Analysis Python - See It in Action

Why transformation reshapes data for analysis

📖 Scenario: Imagine you have sales data for a small store. The data is messy and hard to analyze because it is not organized well. You want to reshape it so it is easier to see total sales per product and per month.

🎯 Goal: You will reshape the sales data from a list of records into a table format that shows total sales for each product by month. This will help you understand which products sell best in which months.

📋 What You'll Learn

Create a list of dictionaries with sales data including product, month, and sales amount

Create a list of months to use as columns

Use a dictionary comprehension to reshape the data into a nested dictionary with products as keys and monthly sales as values

Print the reshaped data to see the total sales per product per month

💡 Why This Matters

🌍 Real World

Data often comes in messy formats. Reshaping it helps us analyze and understand it better, like seeing sales trends over time.

💼 Career

Data analysts and scientists frequently reshape data to prepare it for reports, dashboards, and decision-making.

1

Create the sales data list

Create a list called sales_data with these exact dictionaries: {'product': 'Apples', 'month': 'Jan', 'sales': 150}, {'product': 'Apples', 'month': 'Feb', 'sales': 200}, {'product': 'Bananas', 'month': 'Jan', 'sales': 100}, {'product': 'Bananas', 'month': 'Feb', 'sales': 120}, {'product': 'Cherries', 'month': 'Jan', 'sales': 75}

Data Analysis Python

# Create the sales_data list with the given dictionaries
# Your code here

Hint

Use a list with dictionaries. Each dictionary has keys 'product', 'month', and 'sales' with the exact values given.

2

Create the list of months

Create a list called months with these exact strings: 'Jan' and 'Feb'

Data Analysis Python

sales_data = [
    {'product': 'Apples', 'month': 'Jan', 'sales': 150},
    {'product': 'Apples', 'month': 'Feb', 'sales': 200},
    {'product': 'Bananas', 'month': 'Jan', 'sales': 100},
    {'product': 'Bananas', 'month': 'Feb', 'sales': 120},
    {'product': 'Cherries', 'month': 'Jan', 'sales': 75}
]
# Create the months list with 'Jan' and 'Feb'
# Your code here

Hint

Just create a list with the two month names as strings.

3

Reshape the sales data using dictionary comprehension

Create a dictionary called reshaped_data using dictionary comprehension. For each unique product in sales_data, create a nested dictionary with keys from months and values as total sales for that product and month. If no sales exist for a month, use 0.

Data Analysis Python

sales_data = [
    {'product': 'Apples', 'month': 'Jan', 'sales': 150},
    {'product': 'Apples', 'month': 'Feb', 'sales': 200},
    {'product': 'Bananas', 'month': 'Jan', 'sales': 100},
    {'product': 'Bananas', 'month': 'Feb', 'sales': 120},
    {'product': 'Cherries', 'month': 'Jan', 'sales': 75}
]
months = ['Jan', 'Feb']
# Create reshaped_data dictionary using dictionary comprehension
# Your code here

Hint

Use nested dictionary comprehension. The outer dictionary keys are products. The inner dictionary keys are months. Use sum() to add sales matching product and month.

4

Print the reshaped data

Write a print statement to display the reshaped_data dictionary.

Data Analysis Python

sales_data = [
    {'product': 'Apples', 'month': 'Jan', 'sales': 150},
    {'product': 'Apples', 'month': 'Feb', 'sales': 200},
    {'product': 'Bananas', 'month': 'Jan', 'sales': 100},
    {'product': 'Bananas', 'month': 'Feb', 'sales': 120},
    {'product': 'Cherries', 'month': 'Jan', 'sales': 75}
]
months = ['Jan', 'Feb']

products = set(item['product'] for item in sales_data)
reshaped_data = {
    product: {
        month: sum(item['sales'] for item in sales_data if item['product'] == product and item['month'] == month)
        for month in months
    }
    for product in products
}
# Print the reshaped_data dictionary
# Your code here

Hint

Use print(reshaped_data) to show the final nested dictionary.

Practice

1. Why do we transform data before analysis?

easy

A. To reshape data for easier comparison and visualization

B. To delete important information

C. To make data harder to understand

D. To increase the file size unnecessarily

Solution

1. Step 1: Understand the purpose of data transformation

   Data transformation changes the shape or format of data to make it easier to analyze.

2. Step 2: Identify the benefit of reshaping data

   Reshaping helps in comparing values side-by-side and creating clear visualizations.

3. Final Answer:

   To reshape data for easier comparison and visualization -> Option A

4. Quick Check:

   Transformation = Easier analysis [OK]

Hint: Think: transformation makes data easier to use [OK]

Common Mistakes:

• Confusing transformation with deleting data
• Believing transformation makes data more complex
• Thinking transformation increases file size

2. Which of the following is the correct way to use pandas melt to reshape data?

easy

A. pd.melt(df, id_vars=['score'], value_vars=['id'])

B. pd.melt(df, id_vars=['id'], value_vars=['score'])

C. pd.melt(df, id_vars='score', value_vars='id')

D. pd.melt(df, id_vars=['score', 'id'])

Solution

1. Step 1: Recall pandas melt syntax

   pandas melt requires id_vars as columns to keep and value_vars as columns to unpivot.

2. Step 2: Check pd.melt(df, id_vars=['id'], value_vars=['score'])

   pd.melt(df, id_vars=['id'], value_vars=['score']) correctly uses id_vars=['id'] and value_vars=['score'], matching the syntax.

3. Final Answer:

   pd.melt(df, id_vars=['id'], value_vars=['score']) -> Option B

4. Quick Check:

   Correct melt syntax = pd.melt(df, id_vars=['id'], value_vars=['score']) [OK]

Hint: id_vars are columns to keep, value_vars to unpivot [OK]

Common Mistakes:

• Swapping id_vars and value_vars
• Passing strings instead of lists
• Omitting value_vars parameter

3. Given this DataFrame:

import pandas as pd
df = pd.DataFrame({
  'id': [1, 2],
  'math': [90, 80],
  'english': [85, 88]
})
reshaped = pd.melt(df, id_vars=['id'], value_vars=['math', 'english'])
print(reshaped)

What is the output?

medium

A. A DataFrame with columns: id, math, english and 2 rows

B. A DataFrame with columns: variable, value and 2 rows

C. A DataFrame with columns: id, variable, value and 4 rows

D. An error because of wrong melt usage

Solution

1. Step 1: Understand the original DataFrame

   It has 2 rows and 3 columns: id, math, english.

2. Step 2: Apply pd.melt with id_vars=['id'] and value_vars=['math', 'english']

   This reshapes math and english columns into two rows per id, creating 4 rows total with columns id, variable, value.

3. Final Answer:

   A DataFrame with columns: id, variable, value and 4 rows -> Option C

4. Quick Check:

   melt reshapes wide to long with 4 rows [OK]

Hint: Melt stacks columns into rows, increasing row count [OK]

Common Mistakes:

• Expecting same number of rows after melt
• Confusing columns after melt
• Thinking melt causes error here

4. What is wrong with this code snippet?

import pandas as pd
df = pd.DataFrame({'id':[1,2],'score':[10,20]})
reshaped = pd.melt(df, id_vars='id', value_vars='score')
print(reshaped)

medium

A. print statement is missing parentheses

B. DataFrame creation syntax is incorrect

C. pd.melt does not exist in pandas

D. id_vars and value_vars should be lists, not strings

Solution

1. Step 1: Check pandas melt parameters

   id_vars and value_vars expect list of column names, not single strings.

2. Step 2: Identify error cause

   Passing 'id' and 'score' as strings causes pandas to treat each character as a column, leading to error or wrong output.

3. Final Answer:

   id_vars and value_vars should be lists, not strings -> Option D

4. Quick Check:

   Use lists for id_vars and value_vars [OK]

Hint: Always pass lists for id_vars and value_vars [OK]

Common Mistakes:

• Passing single strings instead of lists
• Misunderstanding melt parameters
• Assuming melt is not in pandas

5. You have sales data in wide format:

import pandas as pd
df = pd.DataFrame({
  'store': ['A', 'B'],
  'Jan': [100, 150],
  'Feb': [120, 160],
  'Mar': [130, 170]
})

You want to analyze monthly sales trends easily. Which transformation helps and why?

hard

A. Use pd.melt to convert months into one column for easier trend analysis

B. Use df.sort_values() to reorder rows by store

C. Use df.drop() to remove month columns

D. Use df.groupby('store') without reshaping

Solution

1. Step 1: Identify the data format and goal

   Data is wide with months as columns; goal is to analyze trends over months.

2. Step 2: Choose transformation to reshape data

   Using pd.melt converts month columns into a single 'month' column and a 'sales' column, making trend analysis easier.

3. Step 3: Explain why other options don't fit

   Sorting or dropping columns doesn't reshape data for trend analysis; grouping without reshaping keeps data wide.

4. Final Answer:

   Use pd.melt to convert months into one column for easier trend analysis -> Option A

5. Quick Check:

   Wide to long format helps trend analysis [OK]

Hint: Melt wide months into one column for trends [OK]

Common Mistakes:

• Trying to analyze trends without reshaping
• Dropping useful columns
• Confusing sorting with reshaping