In some cases, you may need to be able to execute a recipe across multiple instances of identical datasets. For example, if your source dataset is refreshed each week under a parallel directory with a different timestamp, you can create a variable to replace the parts of the file path that change with each refresh. This variable can be modified as needed at job runtime. In Trifacta® Self-Managed Enterprise Edition, parameterization enables you to manage executions of the same recipe steps across serialized datasets the paths to which can be managed via variable.
Suppose you have imported data from a file system source, which has the following source path to weekly transaction logs:
In the above, you can infer a date pattern in the form of
2018/01/29, which suggests that there may be a pattern of paths to transaction files. Based on the pattern, it'd be useful to be able to do the following:
- Import data from parallel paths for other weeks' data.
- Sample across all of the available datasets.
- Execute jobs based on runtime variables that you set for other transaction sets fitting the pattern.
- Pass in parameterized values through API to operationalize the execution of jobs across weeks of transaction data.
The above example implements a Datetime parameter on the path values, creating a dataset with parameters.
You can use the following types of parameters to create datasets with parameters:
- Datetime parameters: Apply parameters to date and time values appearing in source paths.
- When specifying a Datetime parameter, you must also specify a range, which limits the range of the Datetime values.
- Variables: Define variable names and default values for a dataset with parameters. Modify these values at runtime to parameterize execution.
- Pattern parameters:
- Wildcards: Apply wildcards to replace path values.
- Regular Expressions: You can apply regular expressions to specify your dataset matches. Please see the limitations section below for more information.
- Trifacta patterns: The platform supports a simplified means of expressing patterns.
- For more information on Trifacta patterns, see Text Matching.
For more information, see Create Dataset with Parameters.
- You cannot create datasets with parameters from uploaded data.
- You cannot create datasets with parameters from Microsoft Excel files.
- You cannot create dataset with parameters from multiple file types.
- File extensions can be parameterized. Mixing of file types (e.g. TXT and CSV) only works if they are processed in an identical manner, which is rare.
- You cannot create parameters across text and binary file types.
- You cannot apply parameters to write or publishing operations.
- For regular expression patterns, the following reference types are not supported due to the length of time to evaluate:
Backreferences. The following example matches on
cxcyet generates an error:
Lookahead assertions: The following example matches on
a, but only when it is part of an
abparttern. It generates an error:
Creating Dataset with Parameters
From file system
When browsing for data on your default storage layer, you can choose to parameterize elements of the path. Through the Import Data page, you can select elements of the path, apply one of the supported parameter types and then create the dataset with parameters.
NOTE: Matching file path patterns in a large directory can be slow. Where possible, avoid using multiple patterns to match a file pattern or scanning directories with a large number of files. To increase matching speed, avoid wildcards in top-level directories and be as specific as possible with your wildcards and patterns.
For more information, see Create Dataset with Parameters.
From relational source
If you are creating a dataset from a relational source, you can apply parameters to the custom SQL that pulls the data from the source.
NOTE: Avoid using parameters in places in the SQL statement that change the structure of the data. For example, within a
SELECT statement, you should not add parameters between the
For more information, see Create Dataset with SQL.
Managing Datasets with Parameters
Datasets with parameters in your flows
After you have imported a dataset with parameters into your flow:
- You can review any parameters that have been applied to the dataset through the Parameterization in Flow view.
- When the dataset with parameters is selected, you can use the right panel to review and edit the parameters that are applied to it.
- You can change the default value applied to the parameter through the Parameters panel in Flow View.
For more information, see Flow View Page.
Tip: You can review details on the parameters applied to your dataset. See Dataset Details Page.
Sampling from datasets with parameters
When a dataset with parameters is first loaded into the Transformer page, the initial sample is loaded from the first found match in the range of matching datasets.
To work with data that appears in files other than the first match in the dataset, you must create a new sample in the Transformer page. Any sampling operations performed within the Transformer page sample across all matching sources of the dataset.For more information, see Overview of Sampling.
Scheduling for datasets with parameters
Schedules can been applied to a dataset with parameters. When resolving date range rules for scheduling a dataset with parameters, the schedule time is used.
For more information, see Add Schedule Dialog.
Since Trifacta Self-Managed Enterprise Edition never touches the source data, after a source that is matched for a dataset with parameters has been executed, you should consider removing it from the source system or adjusting any applicable ranges on the matching parameters. Otherwise, outdated data may continue to factor into operations on the dataset with parameters.
NOTE: Housekeeping of source data is outside the scope of Trifacta Self-Managed Enterprise Edition. Please contact your IT staff to assist as needed.
Runtime Parameter Overrides
When you choose to run a job on a dataset with parameters from the user interface, any variables are specified using their default values.
Through the Run Job page, you can specify different values to apply to variables for the job.
NOTE: Values applied through the Run Job page to variables override the default values for the current execution of the job. Default values for the next job are not modified.
For more information, see Run Job Page.
In the Job Results page, click View Parameters to view the parameter names and values that were used as part of the job, including the list of matching datasets. See Job Results Page.
You can schedule jobs for datasets with parameters. See Schedule a Job.
Operationalization with Parameters
Through the CLI, you can execute jobs against datasets with parameters. When you download the CLI bundle from the application, the path in the sources file references the first matching dataset among all of the matches. You cannot apply any variable overrides at job runtime.
Tip: External to the platform, you can create a script to manage parameterization. For an example of how to do it, please see CLI Example - Parameterize Job Runs.
For more information, see CLI for Jobs.
By default, parameterization is enabled. For more information on disabling this feature, see Miscellaneous Configuration.
This page has no comments.