Reg2 Template?

[Kevin Blaine]

#18337

Hi Dr. Shah — I was reviewing recent updates to the master asdocx.ado file, and it looks like you’ve recently added support for a Reg2 template. Awesome! I’ve been playing around with it, and I’ve found that it only works part of the time. I thought it might be user error, so I went to look up additional details on how to implement it (and or modify it for my own template), and there isn’t any documentation on the website. In fact, I can’t find any additional details for a number of updates since 2022. Any help would be appreciated!

Thanks,
Kevin

[Kevin Blaine]

#18359

Hi Dr. Shah — just following up on this. Thanks!

[Kevin Blaine]

#18371

Hi again Dr. Shah — any chance you might be able to provide some clarification?

[Attaullah Shah]

#18413

Dear Kevin
Thanks for the follow up and the initial query. I have now added a page for reg2, visit it to get more details about the template.

[Kevin Blaine]

#18444

Hi Dr. Shah — thanks for this! I’m running into an error:

asdocx regress score i.asian, label template(reg2) replace save($Results/Regress.docx)


Error:

i.asian invalid name
r(198);

So, I remove the factor notation and try this:

asdocx regress score asian, label template(reg2) replace save($Results/Regress.docx)

A new error:

variable previous_sheetname undeclared
variable sheets undeclared
 masterclass_populator():  3499  flexmat_to_excel() not found
                 :     -  function returned error [1]
r(3499);

Interestingly, this is the same error I received on the new thread I posted a few minutes ago. Any ideas?

[Attaullah Shah]

#18459

Dear Kevin,

Thank you for your interest in the reg2 template. The reg2 template treats the first variable as the independent variable, followed by a list of dependent variables. This is indeed documented on the reg2 page.

It’s also important to note that Stata does not allow factor variables in the dependent variable. This is a limitation of the software and not the template itself. The reg2 template is designed to work within these constraints.

I hope this clarifies your query. If you have any more questions or need further assistance, feel free to ask.

[Kevin Blaine]

#18465

Thanks Dr. Shah. I understand now. I was thinking it the opposite — that there was one dependent, with multiple independent. So in your example on the help page, each of those ORs represents the odds of the corresponding variable occurring when Robotic==1?

In other words, the result of this Reg2 Template syntax:

asdocx logistic i.Robotic death cdarrest1, vce(cl new_ID) label template(reg2) replace

Is equivalent to the stacked result of this Stata syntax:

logistic death i.Robotic
logistic cdarrest1 i.Robotic

Am I understanding correctly?

Thanks!

[Attaullah Shah]

#18478

Yes, you understood it correctly.

[Kevin Blaine]

#18497

Hi Dr. Shah — I’m loving this template. I wonder, how would I change the decimal placement for ORs, but leave the P-Value decimal place?

For example, I want to report out 2 decimal places for ORs (e.g. 2.06), but leave 3 decimal places for the P-Value (e.g. 0.003)

Thanks!

[Author]

Posts