 |
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: My last word on strange world
I'd like to weigh in with the . detractors.
Logically, a missing value should return FALSE when compared
with any number. Period. (ahem). That StataCorp has chosen
to jettison this logic has caused me no small number of large
headaches, even after years of Stata use. That StataCorp seems
to have jettisoned this logic simply to be consistent with
an arbitrary -sort- implementation is mystifying.
If a user sorts on a variable which has missing values, then
Stata could return an error message saying that the variable
cannot be sorted because of missing values. The user can then
repair or truncate their data so that sorting makes sense.
Just my 2 cents worth.
Jeph
Johannes Geyer wrote:
I'm a bit confused. Why is that '.' has a value in gen, egen and
replace but the maximum value returned by su is not '.'? Surely this is
inconsistent?
Simon
it is not -gen-, -sum- nor -egen- it is the -if- that does not ignore
missing values but interprets them as numbers.
say you want summarize age for all observations that have a wage rate
higher than X and you have missing observations for all that do not report
a wage rate:
sum age if wage > X
then you include them in your condition
Jo
----------------------
Johannes Geyer
Deutsches Institut f�r Wirtschaftsforschung (DIW Berlin)
German Institute for Economic Research
Department of Public Economics
DIW Berlin
Mohrenstra�e 58
10117 Berlin
Tel: +49-30-89789-258
Simon Moore <[email protected]>
Gesendet von: [email protected]
10/01/2008 16:00
Bitte antworten an
[email protected]
An
[email protected]
Kopie
Thema
Re: st: My last word on strange world
I'm a bit confused. Why is that '.' has a value in gen, egen and
replace but the maximum value returned by su is not '.'? Surely this is
inconsistent?
Simon
Austin Nichols wrote:
Allan--
This response I utterly fail to understand--I have seconded your call
for a warning message, several times (is the 3rd time still
seconding?), while pointing out that no software can prevent people
from making errors (hence "some casualties are unavoidable" under any
design). Software that guesses at what you mean is poorly designed,
but software that announces the direct implications of what you coded
is just offering friendly guidance.
I have written a draft version of a command that incorporates warning
messages for -gen- and -replace- and -egen- but such a command will
not protect users who have a condition with implications for a later
-gen- or -replace- or -egen- e.g.
_pctile y if x>42, nq(100)
g hi_x_yp5=r(r5)
Writing a more general warning system would be up to Statacorp, since
it would need to take account of missing values in locals/globals and
scalars as well, which sounds like a major undertaking.
On Jan 10, 2008 6:09 AM, Allan Reese (Cefas) <[email protected]>
wrote:
Austin Nichols asks:
... on the issue of how often a qualifier such as "if x > 42"
can lead to an incorrect analysis, I think there are many dangers of
this sort, and many users fall prey to them. Users of other software
fall under other wheels, and it seems that some casualties are
unavoidable.
Is an attitude I refute utterly and smacks horribly of Harry Lime's
soliloquy on the Riesenrad (big wheel - Third Man):
"If I offered you �20,000 for every dot that stopped - would you
really, old man, tell me to keep my money? Or would you calculate how many
dots you could afford to spare?...Free of Income Tax, old man...".
Also, the Herald of Free Enterprise disaster started with management
who declined to add a safety alarm when directly asked.
Nick has reminded me I made the same suggestion (warning message) four
years ago. As in the case of the Herald, current law allows software
vendors to deny any responsibility for malfunction or consequential loss.
Some time, this position will be challenged and changed.
Allan
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/