system limit exceeded

[Olena Onishchenko]

Dear Stata users

I have panel data of 1000 stocks. For each stock I have about 10 year of data.

I am running the following code:

forvalues year=2004/2015 {
levelsof id if Year==`year', local(idlist)
foreach id of local idlist {
eststo:quietly reg rtr mon tue wed thu fri lag1r lag2r if id == `id' & Year==`year', noconstant
}
}
esttab using d:\Return_regSMALL.csv, append cells("b(fmt(8))")


After the estimate#300 I get an error: "of system limit exceeded you need to drop one or more models".

What do I need to do to avoid this error?

Thanks a lot in advance!

[Carlo Lazzaro]

Olena:
have you already ruled out that your problem doesn't depend on the flavor of Stata you are running (please see help about- )?

[William Lisowski]

The eststo and esttab commands are not official Stata commands; they come from the Stata Journal, as you are asked to explain in the Statalist FAQ.

In help eststo the command is described as "an alternative to official Stata's estimates store." Looking at help estimates for guidance, we see

You can store up to 300 estimation results in memory so that they are at your fingertips.

This suggests your problem is perhaps inherent in the design of eststo, it may depend on the estimates command. It is not clear to me whether for your purposes you need to be using eststo; perhaps you could use eststab, append within your loop. But as I wrote in response to another of your posts, I am not experienced in the use of esttab; you need your problem to come to the attention of someone who is, and the title of this topic may not accomplish that.

[Olena Onishchenko]

Thank you William and Carlo for your responses.

William, I will open a new topic about eststo following all your advices.

Are you aware about other methods of exporting results into excel after my loop? All I need to export is coefficients of Mon and Fri...I would like to learn another method as eststo, esttab causes too much obsticales fro my estimations.

[William Lisowski]

I use postfile to create a file with the results I need from each analysis, then at the end of the loop, read the file that was created and export it to Excel. See help postfile and the full documentation in the Stata Programming Reference Manual. The command was developed to accumulate results from simulations, but it has proven much more useful than that, for exactly the sort of thing you and I are doing.

The statsby command is also recommended by many, but I have yet to use it myself.