[gradsusr] 64-bit erroneous values grads-2.0.1

[David Adams]

Hi Jennifer,
yes, actually these are the values that I was getting on my 64-bit machine, but
the true values actually range from about 190 to 310 (or 19000 to
31000 if not divided by 100).

I will just stick with my painfully slow 32 bit Linux laptop for now.
GrADS produces the correct values in this case.

thanks for you efforts,
Dave





On 4/18/12, Jennifer Adams <jma at cola.iges.org> wrote:
> Oops! I posted a reply to the wrong thread. This is where this response
> belongs…
>
> I got this:
>
> Scanning description file:  test.ctl
> Data file S10216956_201111201830 is open as file 1
> LON set to -100 -25.24
> LAT set to -56 12.52
> LEV set to 1 1
> Time values set: 2010:3:21:17 2010:3:21:17
> E set to 1 1
> ga-> d cls
> Contouring: 5000 to 65000 interval 5000
>
> The display is very noisy but the range of values appears to be correct. I
> will assume the patch to gaio.c for 2-byte data is working correctly for
> your data. If you can build GrADS from source and you want to try to
> implement the patch in a local build, let me know. Otherwise, you'll have to
> wait for our next release.
> --Jennifer
>
>
>
> On Apr 18, 2012, at 11:21 AM, David Adams wrote:
>
>> Hi Jennifer,
>> here is an example binary  file with the appropriate ctl.
>>
>> thanks,
>> Dave
>>
>> On Wed, Apr 18, 2012 at 6:46 AM, Jennifer Adams <jma at cola.iges.org> wrote:
>> I have recently patched the code for reading 2-byte shorts. Could you send
>> me the binary file that goes with your .ctl so I can make sure it is
>> working with the patch?
>> --Jennfier
>>
>>
>> On Apr 17, 2012, at 1:44 PM, David Adams wrote:
>>
>>> Dear Gradsusers,
>>> I am migrating to a 64-bit Linux desktop with Fedora 16.  I have
>>> installed
>>> grads-2.0.1-bin-CentOS5.6-x86_64.tar
>>>
>>> My GOES 12 Channel 4 (binary) data are giving me strange values.I have
>>> tried with byteswapped and without and the values are incorrrect.The
>>> pattern of the incorrect values (brightness temperature) are correct
>>> which I  assume means that the binary values are simply not being read
>>> right.These binary data works fine with the same ctl on my 32-bit  Fedora
>>> Core 10 box and the following version of grads
>>>
>>> grads v2.0.a8
>>> DSET ^S10216956_201110020515
>>> OPTIONS yrev byteswapped
>>> UNDEF  0
>>> TITLE  Temp. Brilho para America do Sul
>>> XDEF 1870 LINEAR  -100.00  0.04
>>> TDEF 1714 LINEAR  -56.00  0.04
>>> ZDEF   01 LEVELS 1 1
>>> TDEF    1 LINEAR 17:00Z21mar2010 30mn
>>> VARS 1 cls    0 -1,40,2,1 'Temp.
>>> ENDVARS
>>> For byteswapped,
>>> a particular value  is 112
>>> for not byteswapped the same value is 65458
>>> The correct value is 28850  (i.e., 288.5K)
>>> Any ideas,
>>> thanks,
>>> Dave
>>> _______________________________________________
>>> gradsusr mailing list
>>> gradsusr at gradsusr.org
>>> http://gradsusr.org/mailman/listinfo/gradsusr
>>
>> --
>> Jennifer M. Adams
>> IGES/COLA
>> 4041 Powder Mill Road, Suite 302
>> Calverton, MD 20705
>> jma at cola.iges.org
>>
>>
>>
>>
>> _______________________________________________
>> gradsusr mailing list
>> gradsusr at gradsusr.org
>> http://gradsusr.org/mailman/listinfo/gradsusr
>>
>>
>> <S10216956_201111201830.gz><test.ctl>_______________________________________________
>> gradsusr mailing list
>> gradsusr at gradsusr.org
>> http://gradsusr.org/mailman/listinfo/gradsusr
>
> --
> Jennifer M. Adams
> IGES/COLA
> 4041 Powder Mill Road, Suite 302
> Calverton, MD 20705
> jma at cola.iges.org
>
>
>
>