Allocating number into groups as per ratios

[Rishab1]

Hi Team , i have a problem where i have groups for each country and jobid , i need to assign candidates (coming as numbers- should be whole numbers cannot be in fractions) to these groups as per the group % ,so in current scenario I have 3 groups with 50%,30%&20% (percent can change) & there can be more or less no of groups , the job should pick up a number like in June it is 3 and allocate this number as a stream of 1's and continue till the sequence stops and if in certain month one groups gets more number in next the next one should get that and there is another table where i have calculated the upper bound for these groups so that when the upper bound is reached it should show as full in that group ... the numbers i have shown in each month in Raw data needs to be divided in these groups.

The logic for allocation should be we would start from group1 and allocate values in stream of 1's and when the count/sequence stops we would from the next group in the next month.

In Raw data the number shown is cumulative i.e needs to be divided in these groups which i have shown in the output.

Requirement is that the end result should be matching the ratios(for all the months) , no need to maintain ratio for each month.

How can i achieve this in talend .?

[Anonymous]

Hi

I think I don't understand your logic well, can you please explain how the expected result came about? take the first raw data for example, why it is 1 for month 1/6/2022? 5 for month 1/7/2022 etc?

Regards

Shong

[Rishab1]

Hi @Shicong Hong​ , this information is coming from 2 different files like(shown in 1st image) and i have put a inner join on jobid & taken all on jobid and created the 2nd image , the job should pick up a number like in June it is 3 and allocate this number as a stream of 1's and continue till the sequence stops and if in certain month one groups gets more number in next the next one should get that and there is another table where i have calculated the upper bound for these groups so that when the upper bound is reached it should show as full in that group , upper bound is calculated and kept in separate table .... After the calculation i need to pivot (that i can do).

Main issue is how to allocate these candidates to groups as per (round robin) and having comparison going with the upper bound to stop is count is equal to that.