Selecting Columns in Pandas Dataframes

When dealing with data manipulation tasks, selecting specific columns becomes necessary. In Pandas, there are various options for selecting columns.

Option 1: Using Column Names

To select columns by their names, simply pass a list of column names as follows:

df1 = df[['a', 'b']]

Option 2: Using Numerical Indices

If the column indices are known, use the iloc function to select them. Note that Python indexing is zero-based.

df1 = df.iloc[:, 0:2]  # Select columns with indices 0 and 1

Alternative Option: Indexing Using Dictionary

For cases where column indices may change, use the following approach:

column_dict = {df.columns.get_loc(c): c for idx, c in enumerate(df.columns)}
df1 = df.iloc[:, list(column_dict.keys())]

Unrecommended Approaches

The following approaches are not recommended as they can lead to errors:

df1 = df['a':'b']  # Slicing column names does not work
df1 = df.ix[:, 'a':'b']  # Deprecated indexing method

Preserving Original Data

Note that selecting columns only creates a view or reference to the original dataframe. If you need an independent copy of the selected columns, use the copy() method:

df1 = df.iloc[:, 0:2].copy()