<>
Is that "unvaluable" or "invaluable"?
And no, it is not my field, and yet, the paper seems pretty interesting even
for an economist/business administrator w/o special knowledge in this
particular area...
HTH
Martin
-----Ursprüngliche Nachricht-----
Von: [email protected]
[mailto:[email protected]] Im Auftrag von Carlo Lazzaro
Gesendet: Samstag, 26. September 2009 20:56
An: [email protected]
Cc: 'Martin Weiss'
Betreff: st: R: AW: R: RE: odd results after insample
Dear Martin,
many thanks for your unvaluable efforts.
However, I am current interested in data reported in Table 3 and 4 of the
article. I have not access to the related data sets and I am figuring out a
way to mimick them and random sampling from the obtained (far fetched)
distributions.
Thanks a lot again, especially for downloading the paper and devoting your
time to think it over?
Out of curiosity: is the economic evaluation of health care programmes one
of your research fields?
Kind Regards,
Carlo
-----Messaggio originale-----
Da: [email protected]
[mailto:[email protected]] Per conto di Martin Weiss
Inviato: sabato 26 settembre 2009 19.51
A: [email protected]
Oggetto: st: AW: R: RE: odd results after insample
<>
So if you wanted the entire figure 1 under Stata -version- 9.2, you would
probably want to install Bobby`s -findit gendist- and then:
*************
clear*
vers 9.2
//lognormal part
capt prog drop myprog
prog def myprog, rclass
vers 9.2
syntax newvarname(numeric max=1), [obs(integer 100) cov(real 2)]
set obs `obs'
loc sd = sqrt(log(`cov'^2+1))
loc mean = log(1000)-.5*`sd'^2
gen `varlist'=exp(invnormal(uniform())*`sd'+`mean')
qui su `varlist'
ret sca mean=r(mean)
ret sca cv=r(sd)/r(mean)
end
loc gra
loc j 1
//for sample size 2000
foreach cv in 0.25 0.5 1 1.5 2{
myprog lognorm`j', obs(2000) cov(`cv')
loc gra `gra' (kdensity lognorm`j' if lognorm`j'<3000)
loc ++j
}
//see the mean and coeff of variation
tabstat _all, stat(mean cv sd)
tw `gra', legend(off) nodraw /*
*/ name(lognormal, replace)
//gamma part
capt prog drop mynewprog
prog def mynewprog, rclass
vers 9.2
syntax newvarname(numeric max=1) [,obs(integer 100) cov(real 2)]
set obs `obs'
loc shape = (`cov')^(-2)
loc scale = 1000/`shape'
gengamma `varlist', alpha(`shape') beta(`scale')
qui su `varlist'
ret sca mean=r(mean)
ret sca cv=r(sd)/r(mean)
end
loc gra
loc j 1
//for sample size 2000
foreach cv in 0.25 0.5 1 1.5 2{
mynewprog gamma`j', obs(2000) cov(`cv')
loc gra `gra' (kdensity gamma`j' if gamma`j'<3000)
loc ++j
}
//see the mean and coeff of variation
tabstat _all, stat(mean cv sd)
tw `gra', legend(off) /* nodraw
*/ name(gamma, replace)
//combine 'em
gr combine lognormal gamma, /*
*/ cols(1)
*************
HTH
Martin
-----Ursprüngliche Nachricht-----
Von: [email protected]
[mailto:[email protected]] Im Auftrag von Carlo Lazzaro
Gesendet: Samstag, 26. September 2009 16:20
An: [email protected]
Cc: 'Martin Weiss'
Betreff: st: R: RE: odd results after insample
Dear Martin,
thanks a lot for your kind reply.
The approach sketched in my previous message follows the one suggested by:
Briggs, A. and Nixon, R. and Dixon, S. and Thompson, S. (2005) Parametric
modelling of cost data: some simulation evidence. Health Economics 14(4):pp.
421-428.
So far, I have been quite successful with other Stata procedures for drawing
random samples from a given distribution (for instance, -simulate-),
including the approach you kindly advice me about.
Unfortunately, I cannot figure out what went wrong with this last do_file.
Thanks a lot again and enjoy your W_E.
Kind Regards,
Carlo
-----Messaggio originale-----
Da: [email protected]
[mailto:[email protected]] Per conto di Martin Weiss
Inviato: sabato 26 settembre 2009 15.21
A: [email protected]
Oggetto: st: RE: odd results after insample
<>
Just out of curiosity: If you want 20 obs per sample, and 2,000 samples,
should that not lead to 40,000 observations overall?
HTH
Martin
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Carlo Lazzaro
Sent: Samstag, 26. September 2009 15:00
To: [email protected]
Subject: st: odd results after insample
Dear Statalisters,
as an alternative to - simulate - , I have written the following do file
(for Stata 9.2/SE) to draw 2000 random samples, 20 observations each, from a
normal distribution:
drop _all
set more off
set obs 2000
obs was 0, now 2000
g double ln_g_20=.
g double ln_sd_g_20=.
set seed 999
qui gen A=5.37 + 1.19*invnorm(uniform()) in 1/972
qui forvalues i = 1(1)2000 {
qui gen ln_20`i'=A
qui generate random`i' = uniform()
qui sort random`i'
qui generate insample`i' = _n <= 20
qui sum ln_20`i' if insample`i' == 1
replace ln_g_20=r(mean) in `i'
replace ln_sd_g_20=r(sd) in `i'
drop ln_20`i'
drop random`i'
drop insample`i'
}
drop A
However, as a result I have obtained 1721 observations instead of the
expected 2000.
sum ln_g_20 ln_sd_g_20
Variable | Obs Mean Std. Dev. Min Max
-------------+--------------------------------------------------------
ln_g_20 | 1271 5.314033 .3800687 3.79247 6.587941
ln_sd_g_20 | 1271 1.101084 .2835007 .0260279 2.161299
Besides, results are even more puzzling when I increase the number of
samples (again 20 observations each), in that I get a different number of
observation for ln_g and ln_sd_g.
Comments are gratefully acknowledged.
Thanks a lot for your kindness and for your time.
Kind Regards,
Carlo
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
