What is the ideal distribution of positives and Negatives to run a Logistic regression

[arunchandra]

Hi ,
Hope you are doing well!...I have a dataset of orders that went for claims (we had to refund money on those orders) along with variables such as time taken to file the claim, order size, number of tickets raised ....I have 4500 orders that had claims in 2019...Can you please let me know how many orders should i choose that had no claims in order to run a logistic regression to determine the probability of determining whether an order would result in a claim....

Thanks,
Arun

 

[tkhunny]

I don't understand the question. Why must you limit the data?