[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Problem with frequencies

From: Ben Pfaff
Subject: Re: Problem with frequencies
Date: Sun, 12 Apr 2009 21:54:37 -0700
User-agent: Gnus/5.11 (Gnus v5.11) Emacs/22.2 (gnu/linux)

michel <address@hidden> writes:

>> Some of the code in FREQUENCIES in the output display code has
>> changed recently, so what version of PSPP is this?  You can give
>> the Git commit number or, failing that, the name of the branch
>> and the date on which you last updated from it.
> It's the master branch from 20090413.

OK, so this is after the changes in question (commit da299bd98
"Allow users to set the precision of output statistics.")

>> Does PES_PNAD2007.sav have a weighting variable?  If so, what is
>> its format?
> What is an "weighting variable"?

It is a numeric variable whose value indicates how many "copies"
of the case should be considered to exist.  If there is no
weighting variable, then every case has a weight of 1.

> If before the frequency I run an "SAMPLE .1", some values appears
> correctly,
> and if I do a "SAMPLE .01" almost everyone appears correctly.
> Seems like the ones who doesn't appears always are represented as five
> asterisks. Maybe if
> the output have more then 5 digits we get this problem. Look at this
> output, with only 10% of the cases.
> The only ones who appears are the smallest values. 

It seems clear that you are right about this, but it is hard to
see how or why it would happen.  When there is a weighting
variable, the column in question is formatted using the weighting
variable's print format, but when there is no weighting variable,
this column is supposed to have an 8-column (not 5-column)

Could you try running the following commands and show me the

        SHOW WEIGHT.
        SHOW FORMAT.
"A computer is a state machine.
 Threads are for people who cant [sic] program state machines."
--Alan Cox

reply via email to

[Prev in Thread] Current Thread [Next in Thread]