After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 551234 - LINEST should optionally ignore incomplete data records
LINEST should optionally ignore incomplete data records
Status: RESOLVED OBSOLETE
Product: Gnumeric
Classification: Applications
Component: Analytics
git master
Other Linux
: Normal enhancement
: ---
Assigned To: Morten Welinder
Jody Goldberg
: 597069 (view as bug list)
Depends on:
Blocks:
 
 
Reported: 2008-09-07 15:09 UTC by Andreas J. Guelzow
Modified: 2018-05-22 13:29 UTC
See Also:
GNOME target: ---
GNOME version: ---


Attachments
screenshot showing the problem (65.43 KB, image/png)
2008-09-09 02:51 UTC, Andreas J. Guelzow
Details

Description Andreas J. Guelzow 2008-09-07 15:09:44 UTC
If a value in the x or y variables range is missing, LINEST considers that value 0 rather than ignoring the data point containg that value. (This is easy to see when one looks at the degree of freedom returned in the 2nd column of the fourht row when stats=TRUE.)
Comment 1 Andreas J. Guelzow 2008-09-09 02:51:44 UTC
Created attachment 118340 [details]
screenshot showing the problem

The attached image shows the value (marked by an arrow) that should be 0 (or for compatibility with XL, should be #N/A).
Comment 2 Andreas J. Guelzow 2008-09-09 02:58:51 UTC
Sorry the previous screen shot should ave been attached to #550933
Comment 3 Andreas J. Guelzow 2009-10-02 16:14:55 UTC
*** Bug 597069 has been marked as a duplicate of this bug. ***
Comment 4 Andreas J. Guelzow 2009-10-02 18:26:50 UTC
Note that I just tried LINEST on Excel 2007, and things seems to have changed. Excel now returns an error in the presence of empty cells. While this is still not great it would at least remove the silent change of empty to 0.
Comment 5 Morten Welinder 2009-10-04 01:07:07 UTC
I can't match our code with the statement "LINEST replaces missing data
with 0".  We seem to ignore strings and empties separately in both
arguments.  That, of course, doesn't make a whole lot of sense.
Comment 6 Andreas J. Guelzow 2009-10-04 01:16:14 UTC
Nope, we are currently having:
COLLECT_IGNORE_STRINGS | COLLECT_IGNORE_BOOLS

Of course just replacing this with 

COLLECT_IGNORE_STRINGS | COLLECT_IGNORE_BOOLS | COLLECT_IGNORE_BLANKS

doesn't quite do it.

I am working on that.
Comment 7 Andreas J. Guelzow 2009-10-04 20:46:04 UTC
The "replace blank cells with 0" issue has been fixed. Of course we should provide the option of excluding all incomplete records.
Comment 8 Morten Welinder 2010-06-11 19:02:47 UTC
The patch to bug 317426 makes us go to "anything but number is error"
which isn't great, but at least consistent.  It should be fairly easy
to improve the situation to "ignore tuples for which any item is blank"
after the patch for 317426 lands.
Comment 9 GNOME Infrastructure Team 2018-05-22 13:29:02 UTC
-- GitLab Migration Automatic Message --

This bug has been migrated to GNOME's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.gnome.org/GNOME/gnumeric/issues/105.