What happen if we increase the tolerance and maximum number of iterations?

[Phuc Nguyen]

When running STATA for panel data (especially for did_imputation),I was suggested by the code sometimes to increase the number of tolerance and iteration, I am wondering how these things affect my regression result in general ?

[Felix Bittmann]

According to the help file, the program uses an iterative algorithm to find a solution. The tolerance measures how close the current solution is to a "good" solution. That means, the algorithm only stops until the current solution is very close to the given cutoff. So if you increase this number, say, from 10^-6 to 10^-3, a larger deviation from the cutoff is acceptable (that is the relative change of coefficient from one iteration to the next). This makes it easier or faster to find a solution. Same with iterations, usually, the algorithm will halt after 100 iterations that do not meet the tolerance. If you set this value to 500, there are simply more tries to get it right. So basically, if your program does not converge you may try to change the values. This could mean that your results become less precise but in reality this probably does not matter much.