Create Extract Form

The create extract form lets you specify a custom data feed from a collection. These data feeds can be open or close ended. This feature automatically tags test and control entities as well as pre and post time periods in the data set.

The first step collects some high-level information before routing you down the Raw or Aggregated paths, depending on your selections.

Step 1: High-level Information

Extract Type: Raw or Aggregated

Date Format: Specify the format of the output date field

Email: Adding an email to notify is useful for large extracts. KL will send an email when data is ready.

Selected Collections: Choose collections to define extract.

Step 2: Creating A Raw Extract

The next step in creating an extract of raw data is to specify what columns you want in the final files.

Click those from the Available Fields column then press Next Step.

Step 3: Define Pre/Post Periods and Controls.

The next step is to specify the pre and post periods as well as the test/control variable and the values for test and control cells.

You need to specify which items you want to track by selecting a filter mapping file type that has the header "Item".

Be warned, for every day in your pre and post period, the application will produce a file - whether or not there is any data for those days.

You can specify for these files to be delivered to a cloud storage location. See Add Delivery functionality of process pipelines.

Pressing save will commence the extract process. The files will be saved as Static Data files and can be searched for in the Static Data Table.

If you have entered an email address then you will get an email once the extract is complete. This email will contain a list of the new file names.

Step 2a: Creating An Aggregated Extract

The flow for creating aggregated files is similar to those for creating Raw data files, except that aggregation rules must be specified.

The first step is to select the fields you want in the aggregated files.

Step 3a: Specify Aggregation Rules

In this step, the fields from the previous selection are displayed except for Price and Quantity because the application knows they are numeric values.

Select combinations of these fields and then press Add.

Select how you want to group the data. Each group will create a different Static Data File.

Select those you require and then press Add to select them.

When you are done press Next Step.

Step 4a: Specify What To "Count"

For each of the aggregation rules you have specified you must now tell KL what you want to be in each file. For example do you want to look at the sum of Quantity or Price or Price * Quantity by Date? Each calculation you select will be performed for each aggregation rule.


So if you select Count Presence of Item then for Date you will get the number of Dates per Date that the item is sold. This will be a 0 or 1 for this aggregation. Had we selected Date * Transaction then this calculation would be the number of transactions that contain the item on that date. See here for more aggregation examples.

Step 5a: Define Pre/Post Periods and Controls.

The final step is to specify the pre/post period. Which field should be used to calculate the test and control cells and which filters contain the test cell values and control cell values.

You must also specify the list of items to extract - this is mandatory even if no item group is created. The items must be in a mapping file with the header "Item". The search in the collection is case sensitive.

Pressing save will begin the extract process. When it is ready an email will be sent to the one(s) entered in the first step. You can then view and process the extracts further via the Static Data Table view.