[Qgis-user] ordering an attribute table

[Michael Spencer]

Thanks Nathan,

I see where you're coming from, but it's not working. The regex selects all
values. I wonder if the problem lies in the data table? The column format
is 'real' and I generated the column using the zonal statistics plugin from
an underlying raster overlain with a regular grid. Null values seem to have
been produced for cells outside the raster, with nan produced where the
raster has no data.

I recoded the raster output from grass to classify the no data values as
-999. How do I go about selecting these? I thought "column" < 0 would work,
but this doesn't select anything.

Another option could be to use the raster tool and vectorise it,
unfortunately this doesn't produce a whole grid as it only outputs integer
values and joins some cells together.


On 27 September 2013 13:44, Nathan Woodrow <madmanwoo at gmail.com> wrote:

> Hey Michael,
>
> Try this expression using the advanced selection feature:
>
> regexp_match('column', '\\d*') = 1
>
> This is using a regex to match column as any number of digits (\\d*).  If
> it's not a digit it will return 0 so you want to only select the ones that
> = 1.
>
> - Nathan
>
>
>
> On Fri, Sep 27, 2013 at 10:16 PM, Michael Spencer <
> spencer.mike.r at gmail.com> wrote:
>
>> Date: Thu, 26 Sep 2013 18:18:10 +0200
>> From: Bernhard Str?bl <bernhard.stroebl at jena.de>
>>
>>
>> Hi,
>>
>> Am 26.09.2013 18:07, schrieb Michael Spencer:
>> > Having a strange problem at the moment.
>> >
>> > If I open a vector attribute table and click on a column heading to sort
>> > the column then this doesn't seem to work. The column contains mainly
>> > numbers with some NA and some NULL strings.
>> >
>> > The sort action groups most together most NA and NULL, and orders the
>> > numbers, but some strings slip through the gaps and appear where they
>> > aren't wanted in the number sequence!
>> >
>> > I'm working in QGIS 2.0.1 on Ubuntu 12.04.
>> >
>> > The reason I want to do this is to delete all records that aren't
>> > numeric, so if there's a neater way to accomplish this that would be
>> great!
>>
>> A neater way would be to select the records in question and then simply
>> delete the selected. Hint select for nulls with
>> <code>"my_field" is null</code>
>>
>> Thanks Bernhard, this was my original choice. Unfortunately I find the
>> select tool a total black box and can't get my head around it. Your
>> suggestion works for entries with 'null' in, but I can't get a response
>> when I want to select 'nan' values. I've tried as:
>>
>> <"field" is nan>
>> <"field" is 'nan'>
>> <"field" = 'nan'>
>>
>> Both with no joy. Is there a crib sheet somewhere with basic options on?
>>
>> Ta,
>> Michael
>>
>> Bernhard
>>
>> >
>> > Ta,
>> > Michael
>>
>> _______________________________________________
>> Qgis-user mailing list
>> Qgis-user at lists.osgeo.org
>> http://lists.osgeo.org/mailman/listinfo/qgis-user
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/qgis-user/attachments/20130927/bb0fe7e6/attachment.html>