[pdal] writers.las unable to convert scaled value

[Rebecca Clarke]

Thanks for all your help and suggestions.
After a short break, I relooked at the issue today and found it to be a
case of mistaken spatial reference system.
The grid and the las data had different srs which meant there were points
falling outside of the moturiki 1953 region.
In effect creating infinte values.



On Thu, Aug 22, 2019 at 11:48 PM Andrew Bell <andrew.bell.ia at gmail.com>
wrote:

> This is a good thought, Thejus, but that issue only comes up when using
> the --json option with PDAL translate.  Also, even if options were getting
> dropped/replaced, changing the scale won't ever get you Inf values.  They
> almost always come from strange reprojections.
>
> If anyone is interested, this is the issue that Thejus was referring to:
> https://github.com/PDAL/PDAL/issues/2538 .  I haven't had a chance to fix
> it yet.
>
> On Thu, Aug 22, 2019 at 7:33 AM thejus kambi <thejus.kambi at gmail.com>
> wrote:
>
>> Hello Andrew,
>>
>> For one of my queries, you had mentioned that the readers and writers
>> can/will be replaced by command line parameters. Do you think something
>> like that is happening?
>> The query was for "translate" command and not pipeline.
>>
>> Warm regards,
>> Thejus
>>
>> On Thu, Aug 22, 2019 at 6:02 AM Andrew Bell <andrew.bell.ia at gmail.com>
>> wrote:
>>
>>>
>>> Hi Rebecca,
>>>
>>> I'm assuming that you can write the file without the reprojection
>>> filter.  If that is indeed the case, then Adam is right and some point is
>>> being reprojected to an infinite Z value (points that are unable to be
>>> transformed are dropped).  I'm not sure what you would like in the end, but
>>> you can write the output to some format you can examine (text, for
>>> example), to determine the points that are problematic.  There could also
>>> be some problem with the proj specification you've provided, but I'm not
>>> familiar enough with those details to be of much help.
>>>
>>> Best,
>>>
>>> On Thu, Aug 22, 2019 at 1:36 AM adam steer <adam.d.steer at gmail.com>
>>> wrote:
>>>
>>>> Hi Rebecca
>>>>
>>>>
>>>>> I ran lasinfo on the las file and get the following result:
>>>>> Z                3696      58127
>>>>>
>>>>
>>>> Ah those are unscaled values - so scaled to metres it would be 36.96
>>>> and 581.27?
>>>>
>>>> What would be the best way to determine if the geoid is introducing any
>>>>> divide-by-zeros?
>>>>>
>>>>> I loaded the gtx into postgis and these are the results from using the
>>>>> ST_SummaryStats:
>>>>>
>>>>> 'count' 'sum' 'mean' 'stddev' 'min' 'max'
>>>>> 14364 | 4215.86899040639 | 0.293502435979281 | 0.0288004424645927 |
>>>>> 0.17399999499321 | 0.49099999666214
>>>>>
>>>>
>>>> That’s a cool way to do it! I wondered if there were null or nodata
>>>> values in the gtx file for some of your las points, but it appears not…
>>>>
>>>> My last easy check idea is to make sure all the las points sit within
>>>> the moturiki 1953 region.
>>>>
>>>> I don’t feel like I’m being super helpful, sorry - just walking down a
>>>> list of things that might produce an ‘-Infinity'….
>>>>
>>>> Regards,
>>>>
>>>> Adam
>>>>
>>>>
>>>>>
>>>>> Cheers
>>>>>
>>>>> Rebecca
>>>>>
>>>>> On Thu, Aug 22, 2019 at 3:50 PM adam steer <adam.d.steer at gmail.com>
>>>>> wrote:
>>>>>
>>>>>> Hi Rebecca
>>>>>>
>>>>>> What are the min and max Z values for your laz file? does it have a
>>>>>> sneaky infinite value in there?
>>>>>>
>>>>>> …and is your geoid grid introducing any divide-by-zeros to create
>>>>>> un-handleable values?
>>>>>>
>>>>>> Cheers
>>>>>>
>>>>>> Adam
>>>>>>
>>>>> _______________________________________________
>>>> pdal mailing list
>>>> pdal at lists.osgeo.org
>>>> https://lists.osgeo.org/mailman/listinfo/pdal
>>>
>>>
>>>
>>> --
>>> Andrew Bell
>>> andrew.bell.ia at gmail.com
>>> _______________________________________________
>>> pdal mailing list
>>> pdal at lists.osgeo.org
>>> https://lists.osgeo.org/mailman/listinfo/pdal
>>
>>
>
> --
> Andrew Bell
> andrew.bell.ia at gmail.com
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/pdal/attachments/20190829/5db74f8c/attachment.html>