For each unique value in a column, a separate column is created. For each row that contains the value in the source column, an indicator value is inserted in the new column.
- This indicator value can be a literal value or the output of a function.
- If no indicator value is generated, a null value is written.
This transform is used to generate indicator columns, which can be used in statistical analysis.
- It evaluates entire cell values for uniqueness. It does not scan for individual elements in Object or Array data.
- If a row in the source column contains a missing value, an indicator value is added in a new
- It is not appropriate for tabulating counts of strings or patterns in a column. See Countpattern Transform.
Optionally, you can specify a default value, which is applied to all non-indicator value cells in the new column.
NOTE: When this transform is applied in the data grid, it only identifies the unique values in the current sample. If there are other unique values in the entire dataset, new columns are not created for them when the transform is executed across the entire dataset.
|Happy Happy Dog|
|Happy Happy Dog||X|
Syntax and Parameters
|values tocols||Y||transform||Name of the transform|
|col||Y||string||Name of source column|
|value||Y||string||String literal, column, or function call that defines the value to use as the indicator value in any newly generated column|
|default||N||string||String literal, column, or function call that defines the value to use to indicate a false match in any newly generated column|
|limit||N||integer (positive)||Maximum number of columns to generate. Default is |
For more information on syntax standards, see Language Documentation Syntax Notes.
Identifies the column to which to apply the transform. You can specify only one column.
|Yes||String (column name)|
valuestocols transform, this parameter specifies the value to insert in each row of a generated column where the column name of the generated column appears in the same row of the source column. This value can be a string literal, a column reference, or a function.
|Yes||String literal, column reference, or function call|
Optionally, this parameter can be used to specify the value to insert in each row of a generated column where the column name of the generated column does not appear in the same row of the source column. This value can be a string literal, a column reference, or a function.
If this parameter is not specified, a missing value is inserted.
|No||String literal, column reference, or function call|
limit parameter defines the maximum number of columns to create from the unique values detected in the source column. If not specified, the limit is
NOTE: Be careful setting this parameter too high. In some cases, the application can run out of memory generating the results, and your results can fail.
|No. Default value is ||Integer (positive)|
Tip: For additional examples, see Common Tasks.
Tip: For additional examples, see Common Tasks.
Example - Basic valuestocols
This dataset contains onboarding milestones for three employees who joined the company at the same time. The milestones were recorded and organized by date as individual items, so it's not easy to verify that all five milestones have been checked off for each employee:
HR Policies Training
|4/4/16||Happy Chandler||Contact Info|
|4/4/16||Bowie Kuhn||Contact Info|
|4/4/16||Bowie Kuhn||Acquire Computer|
|4/4/16||Bud Selig||Product Training|
|4/5/16||Happy Chandler||HR Policies Training|
|4/5/16||Happy Chandler||Acquire Computer|
|4/5/16||Bowie Kuhn||HR Policies Training|
|4/5/16||Bud Selig||HR Policies Training|
|4/5/16||Bud Selig||Contact Info|
|4/6/16||Happy Chandler||Product Training|
The following transform creates columns for each of the values in the
onboardingChecklist column and adds a
yes value where there is a match for the row:
In the generated columns, you can quickly assess whether all three employees have completed an individual onboarding item:
- Bud Selig has not acquired his computer.
- Bowie Kuhn has not had product training.
|4/4/16||Happy Chandler||Contact Info||yes|
|4/4/16||Bowie Kuhn||Contact Info||yes|
|4/4/16||Bowie Kuhn||Acquire Computer||yes|
|4/4/16||Bud Selig||Product Training||yes|
|4/5/16||Happy Chandler||HR Policies Training||yes|
|4/5/16||Happy Chandler||Acquire Computer||yes|
|4/5/16||Bowie Kuhn||HR Policies Training||yes|
|4/5/16||Bud Selig||HR Policies Training||yes|
|4/5/16||Bud Selig||Contact Info||yes|
|4/6/16||Happy Chandler||Product Training||yes|
Example - Magazine subscriptions
flatten- Flatten values in an array into separate rows in the dataset. See Flatten Transform.
valuestocols- Extract unique instances of values into separate columns, with an indicator added to each row where the unique value is found. See Valuestocols Transform.
The following data covers magazine subscriptions for individual customers. Their subscriptions are stored in an array of values. You are interested in who is subscribing to each magazine.
|Anne Aimes||["Little House and Garden","Sporty Pants","Life on the Range"]|
|Barry Barnes||["Sporty Pants","Investing for Tomorrow"]|
|Cindy Compton||["Cakes and Pies","Powerlifting Plus","Running for Days"]|
|Darryl Diaz||["Investing for Tomorrow","Cakes and Pies"]|
When this data is loaded into the Transformer, you might need to apply a
header to it. If it is in CSV format, you might need to apply some
replace transforms to clean up the
Subscriptions column so it looks like the above.
Subscriptions column contains cleanly formatted arrays, the column is re-typed as Array type. You can then apply the
Subscriptioncombination is now written to a separate row. You can use this new data structure to break out instances of magazine subscriptions. Using the following transform, you can add the corresponding
CustIdvalue to the column: Drop the two source columns: Results:
This page has no comments.