Composite schema

#1

Say I want to add breakfast, lunch and dinner every day with random menus

CREATE TABLE meals
(
   meal_type VARCHAR, -- CustomList('breakfast', 'lunch', 'dinner')
   meal_time  TIMESTAMP, -- Date
   menu VARCHAR -- sentences
)

My current strategy case row%3… for the meal type and this + day(row/3) for my date. my actual problem requires about 17 different required events so this is getting tedious.
Is there a sort of composite schema which allows the multiple rows to be created based on a single random field?

#2

If you include all of the dependent data as multiple columns in a single CSV file, you can upload that file into Mockaroo as a dataset, then reference the columns in your schema using the Dataset Column type. When a schema includes multiple columns from the same dataset, the values are always pulled from the same row, so column values will always correspond correctly.

#3

but there is not a 1-1 relationship between meal_type and menu. Say I want to choose a random menu which is a “Breakfast”. In butchered SQL:

SELECT random(menu)
FROM meals
WHERE meal_type = '{meal_type}'

where meal_type is the field holding that data.
Unless I create 3 private fields and use a formula to choose one of them I don’t know how to accomplish this and that does not seem very scale-able. Let me know if there is something wrong with my thought process here.

#4

You might make both meal_type and menu custom lists, then use the “custom” distribution option for menu. Custom distributions allow you to fill out a probability table to weight different combinations. That way you can handle many-to-many correlations.

#5

so break them into multiple datasets?

#6

Yes, or if they are only a few items each you can just use custom list. Custom lists and single column datasets are essentially equivalent.

#7

I was thinking I could get some more functionality if I moved away from the website gui and started using Python+curl, does this seem like a reasonable assessment?