Pigment Champs! Could you please help me understand how Pigment is handling DevOps (Or is it still not available)? Is there any Best Practice document on Pigment (Like Planual in Anaplan)? I see many Pigment sponsored apps have hardcoding in the formula. For Example - As Per Pigment: 1+ INP_YearlySalaryIncrease[BY CONSTANT: Month.Year][FILTER: Month.'Month of Year'= 'Month of Year'."February" ][BY CONSTANT: EE_Data_Stand.Region] I want to write it this way - (1+'01. INP Yearly Salary Increase')[BY CONSTANT:Month.Year][FILTER: Month.'Month of Year'= '00. INP Month of Increase'][BY Constant: 'EE Data Stand'.Region], here the inputs can be changed easily and freely. How does Pigment’s engine perform calculations? How many cores does it leverage at the same time? Is there any recommendation about the source data, that it should have a unique key? I understand that transactional dimension does not need one but it means that you would have to delete and reload/refresh the entire transactional data set as and when there is a change in the existing row as there is no way Pigment can identify which row has been changed due to the change in the source data? Has anyone performed any testing on the load timings when the data is around 10 million records? Also is there any limit on the number of rows that can be added to dimensions? Thanks,Misbah

Hi Misbah !nice to have you here :)what do you mean by devops ?hard coding is a big no no and it is stated in our modeling palette, which I'llbe able to share promptly :)without going into details, it’s based on a cloud computing engine, not limited to a single machine, so it really depends how we scale itUNIQUE ID always. The fact that TL don’t need it is only a shortcut (very handy one)no problem to load huge data sets. I was just chatting about an SA on a project loaded files with 20m recordsfor dimensions themselves, no specific limits. however depending how you use them, you need to be considerate. we Have a customer running most of their calcs on a 1.3m items dimensions, it works. Some 200k+ is common

Solved

Queries related to Pigment

1 month ago
26 July 2024
5 replies
65 views

Pigment Champs!

Could you please help me understand how

Pigment is handling DevOps (Or is it still not available)?
Is there any Best Practice document on Pigment (Like Planual in Anaplan)? I see many Pigment sponsored apps have hardcoding in the formula. For Example - As Per Pigment:
1+ INP_YearlySalaryIncreaseaBY CONSTANT: Month.Year]aFILTER: Month.'Month of Year'= 'Month of Year'."February" ]gBY CONSTANT: EE_Data_Stand.Region]
I want to write it this way - (1+'01. INP Yearly Salary Increase')eBY CONSTANT:Month.Year].FILTER: Month.'Month of Year'= '00. INP Month of Increase']rBY Constant: 'EE Data Stand'.Region], here the inputs can be changed easily and freely.
How does Pigment’s engine perform calculations? How many cores does it leverage at the same time?
Is there any recommendation about the source data, that it should have a unique key? I understand that transactional dimension does not need one but it means that you would have to delete and reload/refresh the entire transactional data set as and when there is a change in the existing row as there is no way Pigment can identify which row has been changed due to the change in the source data?
Has anyone performed any testing on the load timings when the data is around 10 million records? Also is there any limit on the number of rows that can be added to dimensions?

Thanks,

Misbah

5 replies

Userlevel 6

+12

Nathan
Employee
183 replies
1 month ago
26 July 2024
Answer

Hi Misbah !

nice to have you here :)

what do you mean by devops ?
hard coding is a big no no and it is stated in our modeling palette, which I'll be able to share promptly :)
without going into details, it’s based on a cloud computing engine, not limited to a single machine, so it really depends how we scale it
UNIQUE ID always. The fact that TL don’t need it is only a shortcut (very handy one)
no problem to load huge data sets. I was just chatting about an SA on a project loaded files with 20m records
for dimensions themselves, no specific limits. however depending how you use them, you need to be considerate. we Have a customer running most of their calcs on a 1.3m items dimensions, it works. Some 200k+ is common