[GRASS-user] Anything new about the i.class problem?

[Markus Neteler]

On Thu, May 8, 2008 at 11:08 AM, Hernán De Angelis
<dhdeangelis.listas at gmail.com> wrote:
> Hi all,
>
> About one month ago there has been a discussion about a problem with i.class:
> http://lists.osgeo.org/pipermail/grass-user/2008-March/043736.html
>
> I am experiencing exactly the same problems:
>
> - created group and subgroup (with more than 3 images)

fixed.

> - run i.class interactively: i.class can't locate subgroup

fixed yesterday for command line.
Interactive i.class is still broken, see trac #70.

> - run i.class from command line: generated a signature file that
>   cannot be read by i.maxlik

Now that works again.

> in the latter case the resulting signature file

I get for example:

GRASS 6.4.svn (nc_spm_07):~/grassdata/nc_spm_07/landsat/group/mylsat7_2000/subgroup/mylsat7_2000/sig
> cat lsat7_2000_sig
#
#water
517
67.841393 48.589943 37.564796 14.572534 13.218569 114.733078 11.758221 21.897486
1.525194
0.221657 1.347142
0.326066 0.640988 2.428537
0.116158 0.233285 0.371790 0.477774
0.203367 0.219658 0.424782 0.178885 0.884311
-0.032946 -0.007267 0.005329 0.017442 0.015746 0.420543
0.139838 0.069283 0.334514 0.088315 0.219613 -0.064680 1.447250
0.105741 0.246609 0.211119 0.142139 -0.010477 0.133479 -0.022892 3.910005

i.cluster also works.
i.gensigset/i.smap also work.

Remaining issue: fix i.class interactive startup (use command line
startup meanwhile).

Markus