Store separate estimates after -sqreg- for multiple quantiles

[Anton Ivanov]

Hello dear forum members,

My goal is to store estimates after

sqreg y x1-x60, reps (1000) q(.01 .05 .10 .25 .50 .75 .90 .95 .99)

To do so, I am using

estimates store sqreg

However, when I call for estimates stored this way, they are presented in a single column, which is not convenient with 60 regressors and 9 quantiles. E.g.:

----------------------------
                    (1)  
                     y  
----------------------------
q01
                        
x1            0.0000105  
                 (0.49)  

x2                1.817***
                 (4.22)  

x3               0.0197  
                 (0.11)    
----------------------------
q05
                        
x1            0.0000203  
                 (1.40)  

x2                3.042***
                 (7.46)  

x3              -0.0231  
                 (-0.23)

etc.

What is be the correct way to store each set of estimates corresponding a given quantile under a separate name?

Thankfully,
Anton

[Anton Ivanov]

Dear forum members,

I would be glad to provide you with any additional information, in case it is needed to answer my question. I have never come across such issue previously, and therefore seeking your help.

Sincerely,
Anton

[Anton Ivanov]

To address the aforementioned issue, one of the solutions I tried was:

estimates table, keep(q01: q05:) star(0.05 0.01 0.001)

However, it still reports results as a table with a single column (similar to the example provided above).

Another solution I tried involves (1) saving a matrix of coefficients, i.e.:

matrix list e(b)

e(b)[1,585]
             q01:          q01:          q01:          q01:         q01:        
            x1         x2             x3           x4           x5         
y1    -.72351619     7.789e-06     2.1669899    -.02185399    -.10795103     

...
             q05:          q05:          q05:          q05:          q05:        
             x10           x11            x12         x13        x14         
y1     .09693524     .47687667    -5.632e-06     .34947686     .90969925    

etc.

Then (2) saving the coefficient matrix to regular matrix variable using

mat b = e(b)

And then (3) extracting coefficients by row/column into other variables, e.g.:

local coeff1 = b[1,1]

But is there a plausible way to make a table out of these extracted coefficients with each column representing a set of coefficients for a given quantile? Any advise would be appreciated.