Hi,
Yes, the histogram at the moment includes all values below in the
first bin
and above in the last bin. To avoid this use "-l -1000 -u +1000"
before the -H
option which will force the histogram to exclude values outside of
the bin range.
I'll make the next version do this automatically, and will also add
an optional
output as you requested.
All the best,
Mark
On 10 Oct 2005, at 10:34, Helmar Waiczies wrote:
> Hi Mark,
>
> thnx for the binaries, on Fedora Core 3 the binary for Redhat9 ist
> working fine.
> with one exception:
> when i specify a range e.g avwstats++ input -H 2001 -1000 +1000
> the first bin which is written to file is a lot bigger compared to
> the histogram i made with medx
>
> Bin Count Count
> medx medx avwstats++
> -1000 3 66.000000
> -999 0 0.000000
> -998 0 0.000000
> -997 0 0.000000
> -996 0 0.000000
> -995 0 0.000000
> -994 0 0.000000
> -993 0 0.000000
> -992 0 0.000000
> -991 0 0.000000
> -990 0 0.000000
> -989 0 0.000000
> -988 0 0.000000
> -987 1 1.000000
> -986 0 0.000000
> -985 0 0.000000
> -984 0 0.000000
> -983 0 0.000000
> ...
>
> it seems to sum up all bins which are before -1000
>
> an other question is, if its possible to add a switch to the -H
> option that enables/disables
> writing an other column with the selected bins to the output, like
> in the example above
>
> thnx a lot
> Helmar
>
>
> Am Dienstag, den 04.10.2005, 10:59 +0200 schrieb Helmar Waiczies:
>> ive the same problem here
>> <smiley-9.png>
>>
>> compiler too new (gcc3.4.4 and gcc4.0.0)
>>
>> im running on Fedora Core3, but the RedHat9 binaries are usually
>> working
>>
>> thnx a lot for your help
>>
>> Best regards
>> Helmar Waiczies
>>
>> Am Dienstag, den 04.10.2005, 08:30 +0100 schrieb Mark Jenkinson:
>>> Dear Keith, Yes, the existing code does not compile under gcc3.4
>>> and higher. We have a version internally that does, so I will
>>> generate binary versions of avwstats++ in the next couple of days
>>> and then let you know how to get them. All the best, Mark On 3
>>> Oct 2005, at 23:50, Keith Stegbauer wrote: > Dr. Jenkinson, > Our
>>> lab would very much like to use your new version of avwstats++ >
>>> with the histogram feature. We use Cygwin, Redhat 9, Redhat 7,
>>> and > Mac OSX. Unfortunately, we're having trouble compiling on
>>> all these > platforms (except Redhat 7), our compilers are
>>> upgraded to gcc version > 3.4 or higher. As a result, we receive
>>> the error > "/usr/local/fsl/include/newimage/newimagefns.h:582:
>>> error: `vol' > undeclared (first use this function)" in the
>>> compilation of when we > try to make avwstats++. > Do you happen
>>> to have avwstats++ compiled for any of these > platforms? Thanks
>>> in advance for your valuable time. > > > -KC Stegbauer > Senior
>>> Computer Technician > Diagnostic Imaging Sciences Center >
>>> University of Washington > Seattle WA > > > Mark Jenkinson wrote:
>>> > >> Hi, >> >> I thought that a histogram feature in avwstats++
>>> would be useful >> so I've added one. To be able to use it you'll
>>> need to download >> the source code and compile it (follow the
>>> instructions on the >> FSL FAQ for compiling and installing if
>>> you are not sure). >> Alternatively, >> we could compile a
>>> version for you depending on what machine >> you have. Contact me
>>> personally about this if need be. >> >> The new source code can
>>> be found at: >> http://www.fmrib.ox.ac.uk/fsldownloads/patches/
>>> 3.2/avwstats++.cc >> >> The -H option is probably what you want
>>> (where you can specify >> the min and max range explicitly). >>
>>> >> In addition, just a couple of notes on your previous email: >>
>>> >> The avwmaths seems to be working fine. If there aren't any
>>> values >> between 948 and 1000 then the new maximum value (after
>>> zeroing >> everything from 1000 and above) would be 947. It is
>>> the behaviour >> I would expect. >> >> Also, note that cal_min
>>> and cal_max do *not* have to be set to the >> intensity min and
>>> max values in the image. They are intended to >> act as values to
>>> limit the display range only. Hence you cannot >> rely on them to
>>> reflect the full data range. If you use avwstats++ >> with the -R
>>> option then you'll get the full intensity range. >> >> All the
>>> best, >> Mark
|