What do these red xxxxx means during bootstrap?

[Fredrick orwa]

trap replications (1,000)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 50
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 100
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 150
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 200
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 250
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 300
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 350
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 400
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 450
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 500
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 550
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 600
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 650
xxxxxxxxxxxx
Dear team,
I see my program for during bootstrap is showing red xxxx as the bootstrap continues.so I be worried?
below is my code:
use data/ONLYpafdta,clear
* Define a program to calculate PAF
capture program drop calc_paf
program define calc_paf, rclass
***unser5
mi estimate:proportion hiv_status if agecat1==1 & !missing(ipd_outtcome)
local under5prop=round(r(table)[1,2],0.01)
di "`under5prop'"

***calculate relative risk
*use data/MultipleiMPUTATION_inc,clear
*replace mi_m=mi_m-1

*drop mi_m
*ren mi_m _mj
*bys _mj:gen _mi=_n
*gen _mi_miss=1 if _mi_m==0 & missing(hiv_status)
*replace _mi_miss=0 if missing(_mi_miss) & _mi_m==0

*mi import ice, clear
**under5
mi estimate:nbreg ipd_count hiv_status year if agecat1==1 & missing(ipd_outtcome),exp(midyrpopdx) difficult
di exp(r(table)[1,1])
local RR_under5=round(exp(r(table)[1,1]),0.01)
di "`RR_under5'"

// Calculate PAF
local paf =(`under5prop'*(`RR_under5'-1))/(`RR_under5')
return scalar pafunde5 = `paf'
end
**boot strap
bootstrap r(pafunde5), reps(1000) nodrop: calc_paf

[Rich Goldstein]

here is what the help file for -bootstrap- says, " An "x" is displayed if command returns an error or if any value in exp_list is missing."

[Felix Bittmann]

Basically, it means that your program is not working. And apparently you want to use bootstrapping with imputed data, which is not straightforward. Also, the program has some problems. Bootstrapping works on the current dataset and resamples from that, yet you open a new dataset in your program. This part can never be affected by the bootstrap. If you want to combine bootstrapping with imputation, I have written a short guide here: https://www.preprints.org/manuscript/202401.0813/v1

[Fredrick orwa]

Hi Felix.So I can not use mi estimate to get the point estimate once if I want to bootstrap imputed dataset? from your second point under impute+Boot

[Fredrick orwa]

Hi Felix Bittmann can I kindly get the dataset you are imputing its missing values so that I can learn.From your paper.If possible you can send me in my email "[email protected]"

[daniel klein]

The paper links to a GitHub repository. The file Ampute.do creates the dataset.

[Fredrick orwa]

ooh thanks daniel let me check

[Fredrick orwa]

Hi Daniel what is this command fre in line 105 not working in my stata 17

[Fredrick orwa]

it is in line 108 in github script altthough it appears in line105 if i paste in dofile window

[Rich Goldstein]

-fre- is user-written and available from SSC by typing:

ssc install fre

[Fredrick orwa]

I have managed to install it.Thanks Daniel

[Fredrick orwa]

Thanks so much Rich

[Fredrick orwa]

Dear Felix and Daniel,
hope the below code can work now.Its only taking too long running. Have tried to copy your code for second method IMpute+Bootstrap:
****Trying CI for imputed dataset

****calculating PAF and its CI
*use data/ONLYpafdta,clear
* Define a program to calculate PAF


capture program drop calc_paf
program define calc_paf, rclass
use data/ONLYpafdta,clear
*version 17.0
*syntax varlist [,if] [,in]
*args ipdc hiv year agecat ipdoutc pop
***unser5
bsample, cluster(idcode) idcluster(newid)
local j=1
forval fred=1/20{
while `j'<4{
**proportion
proportion hiv_status if agecat1==`j' & !missing(ipd_outtcome) & _mi_m ==`fred'
local prev_`i'_`j'=round(r(table)[1,2],0.0001)
***negative binomial regression
nbreg ipd_count hiv_status year if agecat1==`j' & missing(ipd_outtcome) & _mi_m ==`fred',exp(midyrpopdx) difficult
local prevRR_`i'_`j'=round(r(table)[1,1],0.0001)

}

}
local j=`j'+1
**proportion
local prop_under5=(`prev_1_1'+`prev_2_1'+`prev_3_1'+`pre v_4_1'+`prev_5_1'+`prev_6_1'+`prev_7_1'+`prev_8_1' +`prev_9_1'+`prev_10_1'+`prev_11_1'+`prev_12_1' ///
+`prev_13_1'+`prev_14_1'+`prev_15_1'+`prev_16_1'+` prev_17_1'+`prev_18_1'+`prev_19_1'+`prev_20_1')/20
ren prop_under5 under5prop

local prop_5to14=(`prev_1_2'+`prev_2_2'+`prev_3_2'+`prev _4_2'+`prev_5_2'+`prev_6_2'+`prev_7_2'+`prev_8_2'+ `prev_9_2'+`prev_10_2'+`prev_11_2'+`prev_12_2' ///
+`prev_13_2'+`prev_14_2'+`prev_15_2'+`prev_16_2'+` prev_17_2'+`prev_18_2'+`prev_19_2'+`prev_20_2')/20


local prop_5Above=(`prev_1_3'+`prev_2_3'+`prev_3_3'+`pre v_4_3'+`prev_5_3'+`prev_6_3'+`prev_7_3'+`prev_8_3' +`prev_9_3'+`prev_10_3'+`prev_11_3'+`prev_12_3' ///
+`prev_13_3'+`prev_14_3'+`prev_15_3'+`prev_16_3'+` prev_17_3'+`prev_18_3'+`prev_19_3'+`prev_20_3')/20

***relative risk
local prop_under5RR=exp((`prevRR_1_1'+`prevRR_2_1'+`prev RR_3_1'+`prevRR_4_1'+`prevRR_5_1'+`prevRR_6_1'+`pr evRR_7_1'+`prevRR_8_1'+`prevRR_9_1'+`prevRR_10_1'+ `prevRR_11_1'+`prevRR_12_1' ///
+`prevRR_13_1'+`prevRR_14_1'+`prevRR_15_1'+`prevRR _16_1'+`prevRR_17_1'+`prevRR_18_1'+`prevRR_19_1'+` prevRR_20_1')/20)
ren prop_under5RR RR_under5

local prop_5to14RR=exp((`prevRR_1_2'+`prevRR_2_2'+`prevR R_3_2'+`prevRR_4_2'+`prevRR_5_2'+`prevRR_6_2'+`pre vRR_7_2'+`prevRR_8_2'+`prevRR_9_2'+`prevRR_10_2'+` prevRR_11_2'+`prevRR_12_2' ///
+`prevRR_13_2'+`prevRR_14_2'+`prevRR_15_2'+`prevRR _16_2'+`prevRR_17_2'+`prevRR_18_2'+`prevRR_19_2'+` prevRR_20_2')/20)


local prop_5AboveRR=exp((`prevRR_1_3'+`prevRR_2_3'+`prev RR_3_3'+`prevRR_4_3'+`prevRR_5_3'+`prevRR_6_3'+`pr evRR_7_3'+`prevRR_8_3'+`prevRR_9_3'+`prevRR_10_3'+ `prevRR_11_3'+`prevRR_12_3' ///
+`prevRR_13_3'+`prevRR_14_3'+`prevRR_15_3'+`prevRR _16_3'+`prevRR_17_3'+`prevRR_18_3'+`prevRR_19_3'+` prevRR_20_3')/20)



// Calculate PAF
local paf =(`under5prop'*(`RR_under5'-1))/(`RR_under5')
return scalar pafunde5 = `paf'

***5to14
local pafx =(`prop_5to14'*(`prop_5to14RR'-1))/(`prop_5to14RR')
return scalar pafunde5to14 =`pafx'

***15 and above
local pafxx =(`prop_5Above'*(`prop_5AboveRR'-1))/(`prop_5AboveRR')
return scalar pafunde15plus=`pafxx'

end
**boot strap
*bootstrap r(pafunde5), reps(1000) nodrop: calc_paf
***Amm just so interested in CI
simulate result=r(pafunde5) , reps(1000) seed (123) dots:calc_paf