As mentioned previously, we will start by loading the time dimension. In order to keep the exercise simple, we will populate a dimension with just a few attributes. The following is the DDL for the corresponding table:
CREATE TABLE DIM_TIME ( dateid integer NOT NULL, year integer NOT NULL, month smallint NOT NULL, day smallint NOT NULL, week_day smallint NOT NULL, week_day_desc CHAR(10) NOT NULL, month_desc CHAR(10) NOT NULL, PRIMARY KEY (dateid));
We will create a transformation that generates a dataset with 1,000 days (that is, 1,000 rows) and loads that data into the time dimension:
- Create a transformation.
- From the Input folder, drag and drop a Generate Rows step.
- Use the Generate Rows step to generate a dataset ...