Create data generator for csv

[anchit-chandran]

Following csv values in dummy_sheet_invalid.csv differ from data model. How to map datamodel values to csv values @eatyourpeas ? Particularly null values (will be blank in resulting csv)

True / False fields

LEFT=our data model's value (thus the data generator value which outputs into csv) (enums)
RIGHT = current value of the dummy_sheet_invalid.csv (generally just true / false)

glucose_monitoring

thyroid_treatment_status

gluten_free_diet

psychological_additional_support_status

All YES_NO_UNKNOWN:

smoking_status
Does the patient smoke?

dietician_additional_appointment_offered
Was the patient offered an additional appointment with a paediatric dietitian?

verbose_name
Was the patient using (or trained to use) blood ketone testing equipment at time of visit?

nan becomes float

Pandas will cast nan columns to float. This means resulting csv columns with nans can either:

1. Have blank values, but the existing values are floats so e.g. an enum of [1,2,99] becomes [1.0, 2.0, 99.0]
2. We specify an int value to fill in for nan, so all rows have value

[eatyourpeas]

Pandas is doing something we do not want. The numbers in the choices are the keys and values are the text passed to us by NPDA. These choices are used in the models to store the results. The keys are all integers but null is allowed, and blank in forms is allowed, as some measures are only captured once a year (eg coeliac)