Pandas

Documentation

Subset columns

df[[col1, col2]] returns a DataFrame with only the specified columns.

Select one column

df[col] returns a 1D Series representing the specified column.

df[[col]] returns a 2D DataFrame with one column.

To filter rows based on column values, use boolean indexing:

# Filter rows where 'age' > 30
filtered_df = df[df['age'] > 30]

Note you have to use df['age'] the single bracket notation to access the column for boolean indexing.

If you want selected columns and filter rows at the same time, you can combine them:

# Select 'name' and 'age' columns, and filter rows where 'age' > 30
result = df.loc[df['age'] > 30, ['name', 'age']]

Note that df.loc[rows, columns] allows you to use bollean indexing for rows and specify columns at the same time. The first argument is for row selection, and the second argument is for column selection.

df.iloc[row_index, column_index] is used for integer-based indexing, where you specify the row and column positions. For example, df.iloc[0, 1] will return the value in the first row and second column of the DataFrame.