Key Concepts of Rasgo 0.1

Rasgo uses a few key terms throughout the product.
****Features - Rasgo refers to columns of data that will be used as inputs to models or dashboards as features.
****Collections - Multiple features are used as these inputs and Rasgo allows you to create sets of features that you can work with prior to using them in models and dashboards. Rasgo refers to these sets of features as collections.
****Sources - Rasgo refers to the raw data that has been connected to Rasgo as sources. Features are defined from the data in a source.
****Dimensions - Not all columns of data are features though. Rasgo uses some columns of data called dimensions to join multiple features together in the collections.
****Granularity - Sometimes the column itself could represent multiple different ways a value could be used.
For example, the date/time “2017-01-01 00:00:00”, or Sunday, January 1, 2017 at midnight, could represent
an hour (midnight on that day),
a day (the first of January),
a week (the week starting (or ending) on that day),
a month (January 2017),
a quarter (Q1 of 2017),
a year (2017).
Each of these possibilities should change how features with this dimension are joined together.
In particular, we do not want to naively join features with this date where it represents a week with other features where this same value represents a day.
Rasgo uses granularity to capture this distinction.
Dimensionality appears in several locations within Rasgo and is the key to allowing features to be combined within Rasgo collections.
Visualize Collections from Tableau
Data Source
Last modified 1yr ago