Rename Columns - Video Solutions

Problem Overview: You receive a table-like data structure (typically a DataFrame) and a list of new column names. The task is to rename the existing columns so they match the provided list while preserving the underlying data.

Approach 1: Direct Column Assignment (O(n) time, O(1) extra space)

The most straightforward approach updates the column metadata directly. Iterate through the column list or assign the new column list to the table's column property. In environments like Python DataFrames or similar structures, setting df.columns = new_names immediately replaces the existing column labels. The operation touches each column once, so the time complexity is O(n) where n is the number of columns, while space remains O(1) because only metadata changes. This method is ideal when the number of new names exactly matches the current column count and no conditional logic is required.

Approach 2: Library Rename Utility (O(n) time, O(n) space)

Many data-processing libraries provide a built-in rename utility that accepts a mapping between old and new column names. You construct a dictionary or map where keys represent existing column labels and values represent their replacements, then call the library helper (for example rename()). Internally the library iterates over the column list and applies the mapping. The traversal takes O(n) time and the mapping structure requires O(n) extra space. This approach is preferred when only some columns change or when you want safer transformations with explicit mappings.

Both approaches rely on basic dataframe metadata manipulation rather than complex algorithms. The key operation is iterating across column identifiers and updating their labels. In some implementations, column names are stored in an array-like structure or managed through a hash map for quick lookup during rename operations.

Recommended for interviews: Direct column reassignment is typically the cleanest and most efficient solution when the problem guarantees the new name list matches the column count. It demonstrates you understand how column metadata works internally. The mapping-based rename utility is useful when partial renaming is required and shows familiarity with built-in library features. Interviewers usually expect the O(n) traversal approach because every column must be processed at least once.