Error in running Quantile Regression in Stata with discrete choice variables Urgent help

[apurva ashish tudekar]

quietly eststo All1: qreg2 `y' i.CYLC i.Shanghai i.Princelings i.Military i.gender i.college_dum i.gschool_dum i.mishu_dum i.ethnic_dum i.abroad_dum agez agez2 agez3 i.ladder i.year if `c', quantile (0.5)
It constantly says convergence problem I tried all options, but it isnt happening , this is mandatory for our project, can someone please help as fast as possible

[Nick Cox]

Your urgency doesn't make advice easier. Is this for an assignment?

Please see

https://www.statalist.org/forums/help#noanswer (which is more relevant than the item title implies, as it includes hints on how this kind of question is not easy to answer, even for experts)

and https://www.statalist.org/forums/help#adviceextras #4

On this information my best guess is that you should retreat to a much simpler model that converges and then build up complexity slowly and carefully. How many parameters are you trying to estimate from how many data points? Looks like a truckload to me.

[Joao Santos Silva]

Is the dependent variable binary? Is that why you mentioned discrete choice?

[apurva ashish tudekar]

Is the dependent variable binary? Is that why you mentioned discrete choice?

yes the dependent variable is a dummy so it is binary

[apurva ashish tudekar]

Your urgency doesn't make advice easier. Is this for an assignment?

Please see

https://www.statalist.org/forums/help#noanswer (which is more relevant than the item title implies, as it includes hints on how this kind of question is not easy to answer, even for experts)

and https://www.statalist.org/forums/help#adviceextras #4

On this information my best guess is that you should retreat to a much simpler model that converges and then build up complexity slowly and carefully. How many parameters are you trying to estimate from how many data points? Looks like a truckload to me.

Yes, its for a project where we have to use quantile regression on a paper which used parametric method. its kind of mandatory.

[Joao Santos Silva]

Standard quantile regression only works for continuous dependent variables. You need the maximum score estimator for a binary dependent variable.