Prepare Data

Filter rows, create calculated columns, group and summarize, merge, and check data quality. Your changes automatically carry into EDA and modeling.

Working dataset

Open a CSV in the workspace above.

Changes are saved locally in this browser. Undo keeps the last 20 changes; Restore original discards changes. Do not use sensitive data on shared computers.

1. Filter rows

ColumnConditionCompare with

2. Calculate a column

New column nameSource column ACalculationSource column BConstant (when applicable)

Missing or nonnumeric values produce blank results. Division by zero produces a blank.

Select or Remove Columns

Choose exactly which columns to keep, or choose columns to remove. Changes affect the working dataset, not the original. Undo and Restore Original remain available.

Math Expression Column

Create a numerical column with an expression such as [Revenue] - [Cost] or ([Score1] + [Score2]) / 2. Supports +, -, *, /, ^ and parentheses.

New column name

Expression

Missing or invalid numbers yield blank cells.

Calculated Columns — IF / THEN / ELSE

Make one dummy variable at a time. Example: IF Team equals Red, THEN 1, ELSE 0.

New column nameSource column

Condition Compare with

THEN value

ELSE value

3. Group & summarize

Build a separate summary table without changing the working dataset.

Group by

Summarize column

Statistic

Summaries do not overwrite your observations or affect your models.

4. Merge datasets

Join your working dataset with a second CSV using matching key columns. Only the matching data is added to your workspace; Undo reverses the merge.

Second dataset (only for merging)

Working dataset key

Second dataset key

Join type

No second dataset selected.

Repeated keys in the second file can create extra rows. Blank keys never match. The preview reports matching and unmatched rows before saving.