> ## Documentation Index > Fetch the complete documentation index at: https://docs.rowbase.com/llms.txt > Use this file to discover all available pages before exploring further. # Operations > Transform your data with powerful, non-destructive operations ## Overview Operations are transformations you apply to datasets. They're **non-destructive** - your original data is always preserved, and you can add, remove, or reorder operations at any time. ## Operation Pipeline Operations are applied in sequence, creating a transformation pipeline: ```mermaid theme={null} graph LR A[Raw Data] --> B[Op 1: Filter] B --> C[Op 2: Transform] C --> D[Op 3: Validate] D --> E[Transformed Data] ``` Click any operation in the pipeline to edit it. Drag to reorder. Changes are applied instantly. ## AI Operations ### Use AI Generate AI-powered content based on other column data. Perfect for enrichment, classification, and content generation. ``` Input: product description Output: AI-generated category, summary, or tags ``` ### Detect Headers Use AI to detect column headers in messy data and rename columns automatically. ``` Before: _UNNAMED_0, _UNNAMED_1 After: product_name, price ``` ## Transform Operations ### Filter Rows Keep or exclude rows based on a condition expression. Filter by a single condition: | Field | Operator | Value | | - | - | - | | status | equals | active | Combine conditions with AND/OR: ``` status = "active" AND created_at > "2024-01-01" AND (plan = "pro" OR plan = "enterprise") ``` **Available operators:** * Text: equals, not equals, contains, starts with, ends with, is empty, is not empty * Number: equals, not equals, greater than, less than, between * Date: equals, before, after, between, is empty * Boolean: is true, is false ### Select Columns Keep only specified columns, discarding the rest. ``` Before: name, age, email, phone, address After: name, email ``` ### Rename Columns Change column names for clarity or consistency. ``` "col1" → "customer_name" "col2" → "email_address" ``` * Bulk rename multiple columns at once * Formulas automatically update to use new names ### Drop Column Remove columns you don't need. * Remove single or multiple columns * Removing a column used in formulas will show a warning ### Copy Column Create a copy of a column with a new name. ``` Before: email After: email, email_backup ``` ### Add Literal Column Add a new column with a constant value. ``` New column: "status" Value: "active" ``` ### Formula Compute values using Excel-like formulas with column references. ``` New column: "full_name" Formula: {{first_name}} + " " + {{last_name}} ``` See [Formulas](/features/formulas) for the complete function reference. ## Text Operations ### Trim Whitespace Remove leading and trailing whitespace from text values. ``` Before: " hello " After: "hello" ``` ### Collapse Whitespace Replace multiple consecutive spaces with a single space. ``` Before: "hello world" After: "hello world" ``` ### Normalize Case Convert text to lowercase, uppercase, or title case. ``` Before: "HELLO" After: "hello" ``` ### Strip Prefix / Suffix Remove a specified prefix or suffix from text values. ``` Before: "PREFIX_value" After: "value" ``` ### Replace Values Find and replace text patterns within values. ``` Find: "NY" Replace: "New York" Column: state ``` ### Snake Case Names Convert all column names to snake\_case format. ``` Before: "Product Name", "First Name" After: "product_name", "first_name" ``` ## Parsing Operations ### Parse Number Convert text to numeric values, handling various formats. ``` Before: "1,234.56" After: 1234.56 ``` ### Parse Boolean Convert text like 'yes', 'no', 'true', 'false' to boolean values. ``` Before: "yes" After: true ``` ### Parse Timestamp Parse date/time strings into standardized timestamp format. ``` Before: "01/15/2024" After: "2024-01-15T00:00:00Z" ``` ### Parse Money Extract numeric amounts from currency strings. ``` Before: "$1,234.56" After: 1234.56 ``` ### Parse Percentage Convert percentage strings to decimal values. ``` Before: "50%" After: 0.5 ``` ### Parse Quantity & Unit Split values like '5 mg' into separate quantity and unit columns. ``` Before: "5 mg" After: 5, "mg" (two columns) ``` ## Null Handling ### Empty to Null Convert empty strings to null values. ``` Before: "" After: null ``` ### Null to Default Replace null values with a specified default value. ``` Before: null After: "(default)" ``` ### Coalesce Fields Use the first non-null value from multiple columns. ``` Columns: nickname, first_name, "Unknown" Result: first non-null value ``` ## Merge Operations ### Union Combine rows from multiple tables into a single output. ``` Table A: 100 rows Table B: 50 rows Result: 150 rows (stacked) ``` ### Join Merge two tables by matching rows on common keys. ``` Left table + Right table Join on: customer_id Result: Combined columns where keys match ``` **Join types:** * Left join (keep all left rows) * Inner join (only matching rows) * Full outer join (keep all rows) ### Map Values Replace values using a lookup table (e.g., codes to names). ``` Lookup: "US" → "United States" Before: "US" After: "United States" ``` ## Identity Operations ### Generate Unique ID Create a unique identifier for each row (like a fingerprint for your data). ``` New column: rb_id Value: rb_abc123... ``` ### Apply Corrections Apply manual data corrections and overrides from the corrections table. ## Validation Operations ### Require Fields Ensure specified fields are present and not null in every row. ### Unique Constraint Verify that values in specified columns are unique across all rows. ### Primary Key Constraint Enforce primary key uniqueness and non-null constraints. ### Validate Schema Validate data types and structure against a defined schema. ### Quarantine If Quarantine rows that fail a custom formula-based validation rule. ``` Rule: {{price}} > 0 Result: Rows with zero or negative price are quarantined ``` Quarantined rows are moved to a separate table for review, not deleted. ## Quality Flags Flag rows that match specific quality issues: | Operation | Flags rows where... | | - | - | | Flag Negative | Numeric field contains negative values | | Flag Zero | Numeric field contains zero values | | Flag If Less Than | One column is less than another | | Flag Invalid Units | Units not in allowed list | | Flag Missing Units | Quantity has no associated unit | ## Operation History Every operation is tracked in the history: * View when each operation was added * See who made changes (in team workspaces) * Undo/redo operations * Reset to any point in history ## Best Practices Operations apply in sequence. Filter before validation to ensure you only validate relevant records. Name your operations (e.g., "Remove inactive users") so teammates understand the pipeline. Always check the preview to verify the operation does what you expect. When pattern-based operations aren't enough, use AI operations to handle nuanced transformations.