pandas - Complete Usage of loc and iloc

Data Import

import pandas as pd

from sklearn.datasets import load_boston

# Load the Boston Housing Prices dataset

boston = load_boston()

boston_df = pd.DataFrame(boston.data, columns=boston.feature_names)

boston_df['PRICE'] = boston.target

 

A complete usage of loc

The .loc[] function in Pandas is a powerful tool for data indexing and selection. With the Boston Housing Prices dataset, you can use .loc[] in various ways to perform data manipulations. Here are all possible usages of .loc[] with the dataset:

 

# Usage of loc

# 1. Selecting specific rows and columns by labels

subset_1 = boston_df.loc[[0, 5, 10], ['CRIM', 'RM', 'AGE', 'PRICE']]

# 2. Selecting all rows for specific columns by labels

subset_2 = boston_df.loc[:, ['CRIM', 'RM', 'AGE', 'PRICE']]

# 3. Selecting all rows for a single column by label

subset_3 = boston_df.loc[:, 'PRICE']

# 4. Selecting a single row by label

subset_4 = boston_df.loc[10]

# 5. Selecting rows based on a condition for a specific column

subset_5 = boston_df.loc[boston_df['CRIM'] < 1]

# 6. Selecting rows based on multiple conditions for different columns

subset_6 = boston_df.loc[

    (boston_df['CRIM'] < 1) &

    (boston_df['RM'] > 6) &

    (boston_df['AGE'] < 50)

]

# 7. Selecting specific rows and all columns using a slice for row labels

subset_7 = boston_df.loc[10:15, :]

# 8. Selecting specific rows and all columns using a slice for column labels

subset_8 = boston_df.loc[:, 'CRIM':'AGE']

# 9. Selecting specific rows and specific columns using slices for both row and column labels

subset_9 = boston_df.loc[10:15, 'CRIM':'AGE']

# 10. Modifying values in specific rows and columns using loc[]

boston_df.loc[boston_df['CRIM'] < 1, 'PRICE'] = 50

# 11. Selecting rows and specific columns using a boolean mask

mask = boston_df['CRIM'] < 1

subset_11 = boston_df.loc[mask, ['CRIM', 'RM', 'PRICE']]

# 12. Combining loc[] with query() for complex filtering

subset_12 = boston_df.loc[

    boston_df.query("CRIM < 1 and RM > 6").index,

    ['CRIM', 'RM', 'PRICE']

]

# 13. Selecting rows based on specific index labels

subset_13 = boston_df.loc[[0, 5, 10]]

# 14. Selecting rows based on a condition and specific columns

subset_14 = boston_df.loc[boston_df['CRIM'] < 1, ['CRIM', 'RM', 'PRICE']]

# 15. Modifying a single cell value using loc[]

boston_df.loc[0, 'PRICE'] = 25

# 16. Selecting rows and specific columns based on complex conditions

subset_16 = boston_df.loc[

    (boston_df['CRIM'] < 1) & (boston_df['RM'] > 6),

    ['CRIM', 'RM', 'PRICE']

]

 

A complete usage of iloc

The .iloc[] function in Pandas is used for integer-location based indexing, which means you can use integer positions to select rows and columns from a DataFrame. Here are all possible usages of .iloc[] with the Boston Housing Prices dataset:

# Usage of iloc

# 1. Selecting a single row by integer index

subset_1 = boston_df.iloc[5]

# 2. Selecting multiple rows by integer index

subset_2 = boston_df.iloc[[0, 5, 10]]

# 3. Selecting specific rows and columns by integer index

subset_3 = boston_df.iloc[[0, 5, 10], [0, 5, -1]]

# 4. Selecting all rows for specific columns by integer index

subset_4 = boston_df.iloc[:, [0, 5, -1]]

# 5. Selecting a single cell value by integer index

cell_value = boston_df.iloc[0, 3]

# 6. Modifying a single cell value using iloc[]

boston_df.iloc[0, 3] = 0.5

# 7. Selecting rows and specific columns using integer slices

subset_7 = boston_df.iloc[10:20, 2:6]

# 8. Selecting rows and all columns using integer slices

subset_8 = boston_df.iloc[10:20, :]

# 9. Selecting all rows and specific columns using integer slices

subset_9 = boston_df.iloc[:, 2:6]

# 10. Selecting rows using a boolean mask with iloc[]

mask = (boston_df['CRIM'] < 1) & (boston_df['RM'] > 6)

subset_10 = boston_df.iloc[mask.values]

# 11. Selecting rows and specific columns using integer positions

subset_11 = boston_df.iloc[[0, 5, 10], [0, 5, -1]]

# 12. Modifying values in specific rows and columns using iloc[]

boston_df.iloc[[0, 5, 10], [0, 5, -1]] = 100

# 13. Combining iloc[] with query() for complex filtering

subset_13 = boston_df.iloc[

    boston_df.query("CRIM < 1 and RM > 6").index,

    [0, 5, -1]

]

# 14. Selecting specific columns using integer positions

subset_14 = boston_df.iloc[:, [0, 5, -1]]

# 15. Selecting a single column by integer position

subset_15 = boston_df.iloc[:, 3]

# 16. Modifying a single column using iloc[]

boston_df.iloc[:, 3] = 50

https://ai-fin-tech.tistory.com/entry/Subsetting-Rows-with-Categorical-Variables

Subsetting Rows with Categorical Variables