Medical Procedures Projection based on COVID lockdown decline

I have a 2.7GB csv file with this columns from Brazil’s health sector:

Procedure_Code;
City_Code;
Disease_Code(ICD10);
YearMonth;
Tratment_Type;
Total_Procedures.

You can think five firsts collumns as a key for Total_Procedures value.

The information is already grouped and has 49.466.133 unique values.

Tratment_Type = 3 is ambulatory_public
Tratment_Type = 4 is hospital_public

Tratment_Type = 5 is hospital_private
Tratment_Type = 6 is ambulatory_private

Tratment_Type 3 and 4 we have year 2020. Tratment_Type 4 and 5 we want to project year 2020 based on what happened with Tratment_Type 3 and 4.
Hi Ilya R.,

I have a 2.7GB csv file with this columns from Brazil’s health sector:

Procedure_Code;
City_Code;
Disease_Code(ICD10);
YearMonth;
Tratment_Type;
Total_Procedures.

You can think five firsts collumns as a key for Total_Procedures value.

The information is already grouped and has 49.466.133 unique values.

Tratment_Type = 3 is ambulatory_public
Tratment_Type = 4 is hospital_public

Tratment_Type = 5 is hospital_private
Tratment_Type = 6 is ambulatory_private

Tratment_Type 3 and 4 we have year 2020. Tratment_Type 4 and 5 we want to project year 2020 based on what happened with Tratment_Type 3 and 4.

I will give an example: we can say that procedures related with cancer diagnosis were 50% less in 2020 than they were in 2019 on first semester because COVID.

We can do a projection for type 5 and 6 (private sector) based on what happend with public sector.

All I want is this: A table with same collumns as source with year 2020 for Tratment_Type 5 and 6.