Operations reference

Table Remodeler provides two main categories of operations for working with tabular data files:

Transformations modify tabular data by restructuring, filtering, or augmenting the content. These operations change the data files and can be used to prepare datasets for analysis, convert between formats, or clean up event files.

Summarizations extract information and generate reports without modifying the original data. These operations are useful for quality assurance, understanding dataset structure, and generating analysis-ready summaries.

Common operation concepts

All operations are specified using JSON configuration files that define a list of operations to execute sequentially. Each operation has:

  • operation: The operation name (e.g., “remove_columns”, “summarize_hed_tags”)

  • description: A human-readable description of what this operation does

  • parameters: Operation-specific parameters controlling behavior

Example operation structure:

[{
    "operation": "remove_columns",
    "description": "Remove unnecessary columns from the dataset",
    "parameters": {
        "column_names": ["column1", "column2"],
        "ignore_missing": true
    }
}]

Transformations vs summarizations

Aspect

Transformations

Summarizations

Data modification

Modify the input data files

Do not modify data files

Output

Modified tabular files

Summary reports (JSON/text)

State

Stateless (process one file at a time)

Stateful (accumulate across files)

Common use

Data cleaning, restructuring, format conversion

Quality assurance, validation, dataset understanding

Summary operation parameters

All summarization operations require two standard parameters:

  • summary_name: A unique identifier for this summary instance

  • summary_filename: Base filename for saving the summary (timestamp and extension added automatically)

Optional common parameter:

  • append_timecode: (Default: false) If true, append timestamp to the filename

Operation categories

See also