The Data Table Tab

The Data Table tab displays dataset contents in tabular format, allowing you to explore data using filters and sorting.

See also Getting Started.

Filter Feature

Enter filter expressions to display only rows that match your conditions. Filters also apply to the Statistics tab, which computes statistics on the filtered data only. Graph Builder has its own filter setting and does not use the Data Table filter. Other analysis tabs also use the full, unfiltered data.

Basic Usage

1. Enter an expression in the filter input field at the top of the Data Table tab
2. The filter applies automatically as you type — valid expressions take effect immediately
3. Only rows matching the condition are displayed. If the rows are narrowed down, the number of matching rows (Showing N of M rows) appears below the filter input

If the expression has an error, such as a syntax error, a column name that does not exist, or a value that does not match the column's type, the input field turns red. Hover over the input field to see the error message. The invalid expression is not applied, and the previous valid filter remains in effect.

How to Write Filter Expressions

Filter expressions use SQL-like syntax supporting comparison operators, LIKE, IN, BETWEEN, and more. Keywords (AND, OR, LIKE, IN, etc.) are case-insensitive. Enclose strings in single quotes. To include a single quote inside a string, write it twice, as in name = 'O''Brien'. Enclose a column name in double quotes when it contains spaces or special characters.

Comparison Operators

Numeric literals can be negative, decimal, or in scientific notation, such as temperature > -5 or distance > 5e3. Decimals can omit the integer part, as in .5.

A value you compare against is read as the type of the column it is compared with. Against a numeric column, age = '25' written in single quotes is compared as the number 25; against a boolean column, active = 'true' is compared as true. A value that cannot be read as the column's type, such as age = 'unknown' against a numeric column, turns the input field red.

Date and datetime columns are compared against a date enclosed in single quotes. You can write a date alone as '2024-01-01', include a time as '2024-01-01T09:00:00Z', or add a time zone offset as '2024-01-01T18:00:00+09:00'. The date part also accepts slashes, as in '2024/01/01'. Forms that denote the same instant give the same result.

Pattern Matching Operators

LIKE and ILIKE apply to text columns only. Giving a pattern for a numeric or date column turns the input field red.

Case sensitivity applies to letters, including full-width letters. Hiragana and katakana, as well as full-width and half-width characters, are treated as distinct by both LIKE and ILIKE.

Pattern syntax:

• % represents any string (0 or more characters)
• _ represents any single character

To match % or _ as literal characters, put a backslash before them. MIDAS matches the character right after a backslash as that character itself rather than as a wildcard. To match a backslash contained in a value, write it twice as \\.

Examples:

• name LIKE '%田%' - Contains "田"
• name LIKE '山%' - Starts with "山"
• name LIKE '%郎' - Ends with "郎"
• email LIKE '%@example.com' - Ends with @example.com
• rate LIKE '100\%' - Matches "100%"
• path LIKE 'C:\\Users%' - Starts with "C:\Users"

Logical Operators

Set and Range Operators

Negation Operator

NULL Check and Boolean Values

Column Names with Spaces or Special Characters

Enclose column names in double quotes:

"Body Mass (g)" > 4000

Column names that start with a digit also need double quotes, since they would otherwise be read as numeric literals. Write them as "5K Time" < 25.

Filter Examples

# Numeric condition
body_mass_g > 4000

# String condition (enclose in single quotes)
species = 'Chinstrap'

# Combining multiple conditions
species = 'Adelie' and body_mass_g > 3500

# Excluding missing values
bill_length_mm IS NOT NULL

# Pattern matching (partial match)
island LIKE '%Dream%'

# Matching multiple values
species IN ('Adelie', 'Gentoo')

# Range specification
body_mass_g BETWEEN 3500 AND 4500

# Date condition (enclose in single quotes)
date_col >= '2024-01-01' and date_col < '2025-01-01'

For example, filtering with species IN ('Adelie', 'Gentoo') displays only the matching rows.

Saving Filtered Results as a Dataset

When a filter is active, a Save Filtered Data button appears next to the filter input. Clicking it opens a dialog where you name the filtered result and save it as a derived dataset. The saved dataset opens in a new tab.

Sort Feature

Click the sort button on a column header to sort data by that column.

Single Column Sort

• 1 click: Ascending order
• 2 clicks: Descending order
• 3 clicks: Remove sort

Multi-Column Sort

Hold Ctrl/Cmd while clicking to sort by multiple columns.

• Ctrl/Cmd+click to add a sort condition (ascending)
• Ctrl/Cmd+click the same column again to switch to descending
• Ctrl/Cmd+click once more to remove the sort

During multi-column sort, priority numbers are displayed on sort buttons (e.g., 1, 2). When you remove a sort condition, the priority numbers of the remaining columns are renumbered. Clicking any column header without Ctrl/Cmd clears the multi-column sort and switches to a single-column sort on that column.

Row Selection

Selecting rows in the Data Table synchronizes the selection state with other tabs (Statistics, Graph Builder, etc.).

Selection Methods

• Single row: Click a row
• Range selection: Hold Shift while clicking
• Add to selection: Hold Ctrl/Cmd while clicking
• Clear row selection: Click the Clear rows button in the Selected Rows tab

Selected rows are highlighted on graphs in the Statistics tab and Graph Builder. Row selection does not affect statistics calculations. To restrict which rows statistics are computed on, use a filter. See How Row Selection Works for details on cross-tab synchronization and saving selections as datasets.

Dataset Operations

Viewing Metadata

Click the ⓘ button next to the dataset name to view basic information about the dataset (dataset name, type, row count, column count).

Column Context Menu

Right-click a column header to open the context menu.

• Edit Column Name: Rename the column.
• Edit Scale of Measurement: Change the measurement scale. The measurement scale affects which statistics are shown in the Statistics tab and which graph types are available in Graph Builder.
• Convert Column Types...: Open the type conversion tab. See The Convert Column Types Tab for details.
• Normalize Variants...: Shown for string and enum columns. Opens a tab for fixing variant spellings, such as differences in letter case or whitespace.
• Number Format...: Set the display format for a numeric column (float64 or int64). Choose from presets like fixed decimals, comma grouping, or significant digits. This overrides the default format set in Settings > Display. The format affects how values are displayed in the Data Table tab and in report data tables that include the column; exported data (CSV, TSV, JSON) always uses raw values.
• Display as Link...: Embeds column values into a URL template and displays them as links. Write placeholders in the template as {column_name}. Each placeholder is replaced with that column's value from the same row, URL-encoded. For example, http://example.com/items/{id} embeds each row's id value into the URL. Only URLs with an http or https scheme become links. Links are rendered in projects you created yourself and in projects opened from MDS files with an Official or Trusted signature. In projects that came from an MDS file with an unknown signer, links are shown as plain text to prevent attacks that lure you into opening malicious URLs. For columns with link display configured, this item changes to Edit Link Display..., which opens the dialog with the current URL template already filled in. Use the Remove button in the dialog to clear the link display.

The default number format is set in Settings > Display and applies to data tables, SQL Query Editor preview, and report data tables. You can choose a preset or enter a custom format string using Python-compatible format spec syntax (e.g., ,.2f for comma-separated with 2 decimal places, .4g for 4 significant digits).

Table Menu

Click the menu button (⋮) at the top right of the table to access dataset operations.

• Edit Data: Edit cell values, add rows, or delete rows (Primary datasets only)
• Add to Report: Add the table to a report
• Export: Export data. See Data Export for details
• View SQL Query: View the original SQL query (derived datasets only)
• Materialize: Save query results to the project file (MDS) (derived datasets only). Normally, derived dataset cache is discarded on save, but Materialize preserves it. Useful for computationally expensive small tables.
• Reload Dataset...: Reload data from the original file or URL. Opens a dialog where you select a CSV, TSV, or text file to reload. For datasets loaded from a URL, the Reload from URL button in the dialog fetches the latest data from the original URL. Row count changes are accepted. You can add columns or change the column order, but every column from the original dataset must still be present, matched by name. If an original column is missing (removed or renamed), an error message listing the missing columns is displayed and the dataset is not updated. In that case, import the file as a new dataset instead. For datasets with non-string column types, the new values must also be convertible to those types; if any value cannot be converted, an error message showing the column names and counts is displayed and the dataset is not updated. When you reload a raw dataset, new values that do not fit the detected types are handled by the On Error setting of the converted dataset's column type conversion. Reloading clears all excluded rows and row comments. On successful reload, dependent derived datasets are automatically recalculated. For derived datasets, this item appears as Reload Source Dataset... and reloads the source Primary dataset. If there are multiple source Primary datasets, a dialog opens first to select which one to reload

Data Export

Select Export from the table menu to export the dataset in CSV, TSV, or JSON format. Only rows matching the current filter are exported, in the current sort order. To export all data, clear the filter first. The row number column is not included.

File Name - Specify the output file name. The file is downloaded with the name you enter plus the extension for the selected format, shown to the right of the input. The input starts with a default value that you can edit freely.

File Format - Select the output format.

Encoding - Select the character encoding. Available for CSV and TSV. JSON always uses UTF-8.

When exporting with Shift-JIS or EUC-JP, if the data contains characters that cannot be represented in the selected encoding (such as emoji or certain accented characters), a warning dialog appears. Choose Export anyway to continue, or Cancel to abort the export. To export in UTF-8 instead, cancel and change the Encoding setting. If you continue, unrepresentable characters are replaced with HTML character references (e.g. é).

Add BOM - Adds a byte order mark when UTF-8 is selected. Prevents garbled characters when opening UTF-8 CSV files in Excel.

Include column headers - Outputs column names in the first row. Available for CSV and TSV.

Export selected rows only - Shown when rows are selected. When checked, only the selected rows are exported. Selected rows are exported in the original row number order, not the current sort order.

Missing Values

Missing values are output as empty fields in CSV and TSV formats. In JSON, they are output as null literals.

Row Number Column

The leftmost column displays row numbers. These row numbers indicate the original data order and do not change when filters or sorting are applied.

Project Save State

Filter expressions and sort order are saved to the project file (MDS). When you reopen the project, the previous filter and sort settings are restored. Row selections are not saved.

Next steps

• The Statistics Tab - View statistics for selected columns
• The Graph Builder Tab - Visualize data with graphs
• The SQL Query Editor Tab - Process data with SQL

See also

• Data Preparation and Import - About data types and measurement scales

Also available as a Markdown file.