Decimals and the base level category in factor variables

[Ozer Erdem]

#15407

Hi,
Would it be possible to specifiy decimal lenght for t values and coefficients seperately?
dec(3) rep(t)
With this option, I get everthing with 3 decimals. It is ok for t values but not for coefficients.

Second question,

Even though I use:
nest nofvlabel

I still get empty rows with 0bn.
0bn.High_A

Is it possible to remove numbering which comes with interactions?
1.High_A

For example, if I add c.A##i.High_A, I get in the word document:
0bn.High_A#c.A

1.High_A#c.A

Thank you.

[Dr. Attaullah Shah]

#15409

Hello Ozer Erdem
On your request, I have added option dect() to control the decimal points of standard errors, t-values, or p-value in the nested regressions. I have also fixed the issue with the nofvlabel option. Here is one example where I used the default three decimal places for regression coefficients and two decimal points for standard errors.

sysuse auto, clear
asdocx reg price mpg i.foreign##c.rep78, replace nest 
asdocx reg price mpg rep78 headroom, nest
asdocx reg price mpg rep78 headroom weight , nest 
asdocx reg price mpg rep78 weight , nest dect(2) nofvlabel

You may like to note that asdocx considers all formatting options of the last command used. You may add these options in each line of code, that does not do any harm though.

[Ozer Erdem]

#15410

Amazing 🙂
Thank you.

[Author]

Posts