More PSPP feedback

View: New views
3 Messages — Rating Filter:   Alert me  

More PSPP feedback

by Rob Malpass-3 :: Rate this Message:

| View Threaded | Show Only this Message

Some parts of this message have been removed. Learn more about Nabble's security policy.

Hi all

 

Some improvements I'd like to suggest to psppire - this is all based on the Windows executable.   [See other post - I can't make the Ubuntu version work at all at the moment]

 

1) The file requester for Import Delimited Text Data is very odd.   It doesn't give the chance to select .csv, .txt or whatever - I think it would be better if it did.

2) In the data editor itself, you can't navigate easily.   For example the following keyboard shortcuts would be very useful:
            PageDn - to go down one "screen full" of data

            Ctrl+Down Arrow to move to the last case

3) When running a simple linear regression, I couldn't see anywhere in the output where it lists the number of observations it has included - I think this should be included somewhere.

4) Does pspp support a Chow test for splitting a population into two groups, running a regression on each and testing for statistically significant differences in the coeffs etc?

 

Thanks

Rob

 


_______________________________________________
Pspp-users mailing list
Pspp-users@...
https://lists.gnu.org/mailman/listinfo/pspp-users

Re: More PSPP feedback

by Ben Pfaff :: Rate this Message:

| View Threaded | Show Only this Message

"Rob Malpass" <rob@...> writes:

> 1) The file requester for Import Delimited Text Data is very odd.   It
> doesn't give the chance to select .csv, .txt or whatever - I think it would
> be better if it did.

I filed this as bug #36054:
        https://savannah.gnu.org/bugs/index.php?36054
You should be CCed.

> 2) In the data editor itself, you can't navigate easily.   For example the
> following keyboard shortcuts would be very useful:
>             PageDn - to go down one "screen full" of data
>
>             Ctrl+Down Arrow to move to the last case

We've got a rewrite of the data editor in the works.
--
Ben Pfaff
http://benpfaff.org

_______________________________________________
Pspp-users mailing list
Pspp-users@...
https://lists.gnu.org/mailman/listinfo/pspp-users

Chow test

by John Darrington-4 :: Rate this Message:

| View Threaded | Show Only this Message

On Sat, Mar 31, 2012 at 02:23:11PM +0100, Rob Malpass wrote:
     
     
     4) Does pspp support a Chow test for splitting a population into two groups,
     running a regression on each and testing for statistically significant
     differences in the coeffs etc?

Not explicitly.  

However, it's easy to split the data set according to some predicate
(use SPLIT FILE).  Then  run your REGRESSION command.

Then you'll have to calculate F (either manually or using PSPP's
COMPUTE command, and look up the significance (in tables, or by
PSPPs CDF.F function).

J'


--
PGP Public key ID: 1024D/2DE827B3
fingerprint = 8797 A26D 0854 2EAB 0285  A290 8A67 719C 2DE8 27B3
See http://keys.gnupg.net or any PGP keyserver for public key.



_______________________________________________
Pspp-users mailing list
Pspp-users@...
https://lists.gnu.org/mailman/listinfo/pspp-users

signature.asc (196 bytes) Download Attachment