Failure to store regression data (b, pvalue, CI) in a new variable in the full database

[Andressa Freire]

Hi,

I got help from the forum to solve my problem and it was great ("How to store p-value and CI in a new variable in beaches-winsten regression?"). I managed to carry out my analyses, but now in other banks I can't reproduce it completely.

The command runs to some extent (as in the example generated by dataex). I believe that the error can occur in some groups of code_obge because the values over time are similar or even the same.
How do I make it run full and "ignore" regressions that give error?

** version Stata 17

Code used:

xtset codigo_ibge ano

levelsof codigo_ibge, local(ids)
gen b = .
gen ll = .
gen ul = .
gen pvalue = .

foreach i of local ids {
capture prais logprev_iac_ ano if codigo_ibge == `i'
if c(rc) == 0 {
matrix M = r(table)
foreach x in b ll ul pvalue {
quietly replace `x' = M["`x'", "ano"] if codigo_ibge == `i'
}
}
else if inlist(c(rc), 2000, 2001) {
continue
}
else {
display as error "Unexpected error: codigo_ibge == `i'"
exit(c(rc))
}
}

gen apc_iac= (-1+(10^ b))*100
gen apc_min_iac= (-1+(10^ll))*100
gen apc_max_iac= (-1+(10^ul))*100





----------------------- copy starting from the next line -----------------------

* Example generated by -dataex-. To install: ssc install dataex
clear
input long codigo_ibge int ano float prev_iac_ byte porte_municipio float(logprev_iac_ b ll ul pvalue apc_iac apc_min_iac apc_max_iac)
316245 2015       100 0         2   -.008813011 -.027603345  .00997732 .2323614    -2.0088198  -6.15813 2.3239558
316245 2016  91.66666 0 1.9622114   -.008813011 -.027603345  .00997732 .2323614    -2.0088198  -6.15813 2.3239558
316245 2017      87.5 0  1.942008   -.008813011 -.027603345  .00997732 .2323614    -2.0088198  -6.15813 2.3239558
316245 2018  91.30434 0 1.9604914   -.008813011 -.027603345  .00997732 .2323614    -2.0088198  -6.15813 2.3239558
316245 2019  90.47619 0 1.9565343   -.008813011 -.027603345  .00997732 .2323614    -2.0088198  -6.15813 2.3239558
316447 2015       100 0         2 -2.458474e-14 -.016400341 .016400341        1 -5.662137e-12 -3.705904  3.848527
316447 2016       100 0         2 -2.458474e-14 -.016400341 .016400341        1 -5.662137e-12 -3.705904  3.848527
316447 2017  91.66666 0 1.9622114 -2.458474e-14 -.016400341 .016400341        1 -5.662137e-12 -3.705904  3.848527
316447 2018       100 0         2 -2.458474e-14 -.016400341 .016400341        1 -5.662137e-12 -3.705904  3.848527
316447 2019       100 0         2 -2.458474e-14 -.016400341 .016400341        1 -5.662137e-12 -3.705904  3.848527
316470 2015  94.11765 2  1.973671             .           .          .        .             .         .         .
316470 2016   94.3662 2 1.9748164             .           .          .        .             .         .         .
316470 2017  95.06173 2 1.9780058             .           .          .        .             .         .         .
316470 2018  95.52238 2  1.980105             .           .          .        .             .         .         .
316470 2019      87.5 2  1.942008             .           .          .        .             .         .         .
316520 2015       100 0         2             .           .          .        .             .         .         .
316520 2016        80 0   1.90309             .           .          .        .             .         .         .
316520 2017        80 0   1.90309             .           .          .        .             .         .         .
316520 2018  95.83334 0 1.9815166             .           .          .        .             .         .         .
316520 2019  90.90909 0 1.9586073             .           .          .        .             .         .         .
316557 2015 66.666664 0 1.8239087             .           .          .        .             .         .         .
316557 2016       100 0         2             .           .          .        .             .         .         .
316557 2017        90 0 1.9542425             .           .          .        .             .         .         .
316557 2018  90.90909 0 1.9586073             .           .          .        .             .         .         .
316557 2019 66.666664 0 1.8239087             .           .          .        .             .         .         .
316680 2015       100 0         2             .           .          .        .             .         .         .
316680 2016  96.15385 0 1.9829667             .           .          .        .             .         .         .
316680 2017      87.5 0  1.942008             .           .          .        .             .         .         .
316680 2018       100 0         2             .           .          .        .             .         .         .
316680 2019  92.85714 0 1.9678153             .           .          .        .             .         .         .
end
label values porte_municipio porte_municipio
label def porte_municipio 0 "pequeno I", modify
label def porte_municipio 2 "medio", modify

------------------ copy up to and including the previous line ------------------

[William Lisowski]

Perhaps the suggested changes in red will start you in a useful direction.

foreach i of local ids {
    capture prais logprev_iac_ ano if codigo_ibge == `i'
    if c(rc) == 0 {
        matrix M = r(table)
        foreach x in b ll ul pvalue {
            quietly replace `x' = M["`x'", "ano"] if codigo_ibge == `i'
        }
    }
    else if inlist(c(rc), 2000, 2001) {
        continue
    }
    else {
        display as error "Unexpected error: codigo_ibge == `i'  error code"  c(rc)
    }
}

[Andressa Freire]

Hi, thank you so much William!

It worked really well!