Data Editor
A spreadsheet-style data preprocessing workbench. Upload CSV, Excel, or JSON files, clean and transform your data, run queries and pivot tables, then export — all without writing code.
Overview
Spreadsheet-style editor
Double-click any cell to edit inline. Select rows or columns to apply bulk operations.
One-click transforms
Fill missing values, apply log/z-score/min-max transforms, encode categoricals — all from the sidebar.
Built-in query engine
GROUP BY, PIVOT, JOIN, FILTER, SORT without writing SQL. Results open as a new tab.
What you can do
How data flows
Getting started
The Data Editor after a file is loaded
Click Upload Files in the sidebar or drag and drop a file anywhere on the page. Multiple files open as separate tabs.
Click Sample Data to load a 1,000-row employee dataset — useful for exploring features without your own file.
Your data appears in the main table. Column stats (type, null count) show below each header when Show stats is on.
The left sidebar organizes all operations — Import/Export, Rows, Columns, Data Quality, Analysis, and Advanced.
Supported file formats
Data table
The main table supports inline editing and bulk selection. Null values are highlighted in rose, and the table previews up to 1,000 rows.
Rows & Columns sidebar
History — undo / redo
Edit cell
Double-click any cell to enter edit mode. Press Enter to confirm, Esc to cancel.
Select columns
Click the checkbox in a column header to select it. Multiple columns can be selected for bulk operations.
Select rows
Click the row checkbox to select. Use the header checkbox to select all visible rows.
Sort
Click the ⋮ menu in any column header → Sort ascending or Sort descending.
Rename column
Click the column name in the header and type a new name directly. Or use Bulk Rename (⋮ menu → Bulk Rename) to rename all columns at once with presets: trim, lowercase, snake_case, camelCase.
Filter rows
Use the Filters button in the toolbar. Stack multiple conditions — all are combined with AND.
Search
The search box above the table filters across all columns instantly.
Bulk Rename Columns dialog
Data quality
Select one or more columns in the table, then use the Data Quality section in the sidebar.
Data Quality sidebar
Fill Missing Values
How to use
- ①Click the checkbox in the column header to select the column(s) with missing values.
- ②In the left sidebar, open the Data Quality section → Fill Missing Values.
- ③Choose a fill method from the dropdown (see table below).
- ④Click Fill. The null cells are replaced immediately. Use Ctrl+Z to undo.
Tip: Use the Missing Heatmap (Analysis section) first to see which columns need attention.
Select columns → choose a method → click the Fill button.
| Method | Description |
|---|---|
| mean | Replace nulls with the column mean (numeric columns) |
| median | Replace nulls with the column median (numeric columns) |
| mode | Replace nulls with the most frequent value |
| zero | Replace nulls with 0 (numeric) or empty string (text) |
| forward | Propagate the last valid value downward |
| backward | Propagate the next valid value upward |
Numeric Transform
How to use
- ①Select one or more numeric columns using the column header checkboxes.
- ②In the sidebar, open Data Quality → Numeric Transform.
- ③Choose a transform from the dropdown.
- ④Click ⚡. A new column is added with the suffix _log, _zscore, etc. The original column is kept.
When to use each transform:
• log / log10 — skewed income, price, or count data before regression
• zscore — comparing variables on different scales, or before PCA/clustering
• minmax — normalizing inputs for neural networks or KNN
• sqrt — count data with moderate skew (less aggressive than log)
Select numeric columns → choose a transform → click ⚡.
Encoding
How to use
- ①Select the categorical column(s) to encode.
- ②In the sidebar, open Data Quality → Encoding → choose an encoding type.
- ③Click Apply. New columns are added; the original column is kept by default.
- ④Drop the original column via the ⋮ menu if it's no longer needed before analysis.
Which encoding to pick:
• One-Hot (OHE) — default choice for linear models, logistic regression, SVM
• Label (LE) — decision trees, random forests only
• Ordinal (OE) — when categories have a clear order (Low → Med → High)
• Target (TE) — high-cardinality columns, but split data first to avoid leakage
• Frequency (FE) — city, ZIP code, or any column with many unique values
Select categorical columns → click Encoding (7 types).
One-Hot
Category → binary columns. Best for unordered categories in linear models. Limited to 50 unique values.
Label
Category → integer (alphabetical). Suitable for tree models only.
Ordinal
Category → integer by a rank you specify. Use when categories have a natural order (S < M < L).
Target
Category → target column mean. Powerful but apply after train/test split to avoid data leakage.
Frequency
Category → frequency ratio (0–1). Good for high-cardinality columns.
Count
Category → occurrence count (integer).
Binary
Category → log₂(N) binary columns. More compact than One-Hot for high-cardinality columns.
Outlier Detection
Select a numeric column → sidebar Advanced section → Outlier Detection.
IQR method
Flags values outside Q1 − k·IQR and Q3 + k·IQR. Default k = 1.5. Robust to extreme outliers.
Z-score method
Flags values where |z| > threshold. Default threshold = 3. Assumes roughly normal distribution.
Actions
Analysis tools
Five built-in tools accessible from the Analysis and Advanced sections of the sidebar.
How to open
In the left sidebar, scroll down to the Analysis section (Query Engine, Pivot Table, Column Profiling, Missing Heatmap) or the Advanced section (Derived Column, Outlier Detection). Each tool opens as a panel on the right side of the screen. Results from Query Engine and Pivot Table open as a new tab.
Analysis sidebar
Query Engine panel
Query Engine
Visual query builder — no SQL required. Open from sidebar → Analysis → Query Engine.
What it does
- •FILTER — keep rows matching a condition (==, !=, >, contains, is null, …)
- •GROUP BY — aggregate by one or more columns (count, sum, mean, min, max, median, std, nunique)
- •PIVOT — reshape rows into columns with an aggregation
- •JOIN — merge with another open tab on a key column (INNER, LEFT, RIGHT, FULL)
- •SORT — order rows by any column ascending or descending
- •SELECT — keep only the columns you need
- •LIMIT / OFFSET — slice a fixed number of rows
- •Results open as a new tab — original data is untouched
- •Example: GROUP BY department → mean(salary) gives average salary per department in one click
- •Example: JOIN sales tab with customers tab on customer_id to enrich your dataset
Pivot Table panel
Pivot Table
Summarize a column by group with count, sum, mean, min, or max. Open from sidebar → Analysis → Pivot Table.
What it does
- •Choose a Group By column and an aggregation function
- •Results show count, % of total, and the aggregated value per group
- •Click any column header to sort
- •Search box to filter groups by name
- •Example: Group By = region, Aggregate = sum(revenue) → instant revenue breakdown by region
Column Profiling panel
Column Profiling
Statistical summary and distribution for any column. Open from sidebar → Analysis → Profile Column.
What it does
- •Numeric columns: min, max, mean, median, std, Q1, Q3, IQR, skewness, kurtosis, outlier count
- •Categorical columns: unique count, cardinality, mode, mode frequency
- •Completeness bar — % of non-null values
- •Distribution histogram (numeric) or value frequency bars (categorical)
- •Data quality score (0–100) with automatic recommendations
- •Use this before any analysis to check for skewness, outliers, or low-quality columns
Missing Heatmap panel
Missing Heatmap
Visual overview of null values across all columns. Open from sidebar → Analysis → Missing Heatmap.
What it does
- •Columns sorted by missing rate (highest first)
- •Bar color: green = 0%, amber = <20%, rose = ≥20%
- •Summary KPIs: total missing cells, clean columns, problem columns
- •Use this as a first step after uploading to decide which columns need Fill Missing
Derived Column panel
Derived Column
Add a new column computed from a formula. Open from sidebar → Advanced → Derived Column.
What it does
- •Builder mode — click columns, operators, and functions to build a formula visually
- •Raw mode — type any JavaScript expression
- •Preview shows the result for the first 5 rows before you commit
- •Supported functions: round(), abs(), log(), sqrt(), floor(), ceil()
- •Example: salary * 1.1 → adds a column with a 10% salary increase
- •Example: age > 30 ? "senior" : "junior" → adds a string classification column
- •Example: (revenue - cost) / revenue → adds a profit margin column
Export
Click Download CSV in the sidebar at any point. Use the dropdown arrow to switch format.
CSV
UTF-8 BOM encoded, compatible with Excel and most tools
Excel
.xlsx with data on the first sheet
JSON
Array of objects — one object per row, keys = column names
Keyboard shortcuts
Press ? anywhere on the page to open the shortcuts panel.
| Shortcut | Action |
|---|---|
Ctrl+Z | Undo |
Ctrl+Shift+Z | Redo |
Ctrl+S | Download as CSV |
Delete | Delete selected rows |
Esc | Clear selection / cancel edit |
Double-click | Edit cell |
Enter | Confirm cell edit |
? | Show keyboard shortcuts |