3.1 Read Data#

From csv file or dict, list, tuple, ndarray, Series objects in Python.

import pandas as pd

# Read from csv file
df = pd.read_csv('data.csv')

# Create DataFrame from Python objects
df_dict = pd.DataFrame({'Name': ['Tom', 'Jerry'], 'Age': [20, 21]}) 
df_list = pd.DataFrame([['Tom', 20], ['Jerry', 21]], columns=['Name', 'Age'])
df_tuple = pd.DataFrame((('Tom', 20), ('Jerry', 21)), columns=['Name', 'Age'])

import numpy as np # ndarray example
data = np.array([['Tom', 20], ['Jerry', 21]])
df_ndarray = pd.DataFrame(data, columns=['Name', 'Age'])

3.2 Select Data#

• Select Columns: Use column names directly or df[['col1', 'col2']].

• Select Rows: Use slicing or conditions, e.g., df[1:3] or df[df['Age'] > 20].

• Select Cells: Combine row and column selection, e.g., df.loc[rows, cols] or df.iloc[row_indices, col_indices].

3.2 Select Data#

# Select a single column
names = df['Name']

# Select multiple columns
subset = df[['Name', 'Age']]

# Select rows by slicing
rows_slice = df[0:2]

# Select rows by condition
adults = df[df['Age'] >= 18]

# Select specific cells
cell_value = df.loc[1, 'Name']  # Second row, Name column

3.3 Filter Data#

# Simple condition
adults = df[df['Age'] >= 18]

# Multiple conditions
young_adults = df[(df['Age'] >= 18) & (df['Age'] <= 30)]

# Using query method
seniors = df.query('Age > 60')

# Filtering with isin method for categorical data
students = df[df['Occupation'].isin(['Student'])]

3.4 Sort Data#

Sorting data is crucial for analyzing datasets where the order of rows is significant.

# Sort DataFrame by a single column
df_sorted = df.sort_values(by='Age')

# Sort DataFrame by multiple columns
df_sorted_multi = df.sort_values(by=['LastName', 'FirstName'])

# Sort in descending order
df_sorted_desc = df.sort_values(by='Age', ascending=False)

3.5 Aggregating Data#

# Group by a single column and calculate mean
grouped_single = df.groupby('Department').mean()

# Group by multiple columns
grouped_multi = df.groupby(['Department', 'Position']).sum()

# Applying multiple aggregation functions at once
grouped_agg = df.groupby('Department').agg({'Salary': ['mean', 'min', 'max']})

3.5.2 Aggregating Data with Pivot Tables#

# The `pivot_table` method is used to generate a Pandas pivot table
import numpy as np

babynames.pivot_table(
    index = "Year",
    columns = "Sex",    
    values = "Count", 
    aggfunc = np.sum, 
)

3.6 Join Data#

• left and right parameters are used to specify the DataFrames to be joined.
• left_on and right_on parameters are for aligning the key columns.

# Merge two DataFrames on a key
merged_df = pd.merge(df1, df2, on='EmployeeID')

# Merge using different keys for each DataFrame
merged_diff_keys = pd.merge(
    df1, 
    df2, 
    left_on='EmployeeID_df1',
    right_on='EmployeeID_df2'
)

3.7 Modify Data#

# Change a value directly
df.loc[0, 'Age'] = 22

# Add a new column
df['Country'] = 'Unknown'  # Adds a new column with all values set to 'Unknown'

# Apply a function to transform data
df['Age in 5 Years'] = df['Age'] + 5

# Using apply() to apply a custom function to each row
df['Name Length'] = df['Name'].apply(len)

3.8 Delete Data#

# Drop a column
df.drop('Age in 5 Years', axis=1, inplace=True)

# Drop a row
df.drop(index=0, inplace=True)

# Drop rows based on condition
df = df[df['Age'] >= 18]

# Set a cell value to NaN
import numpy as np
df.loc[df['Age'] < 18, 'Age'] = np.nan