[R-sig-Geo] Implementing a rolling window for stars object

[Edzer Pebesma]

On 10/26/19 4:43 AM, Micha Silver wrote:
> 
> On 25/10/2019 15:55, Edzer Pebesma wrote:
>> Thanks for reporting Micha, and creating a reprex on github Andy; the
>> problem seems to be fixed, now.
> 
> 
> Great, thanks for the quick response.
> 
> I guess reinstalling stars from CRAN should get the fix?

Sorry, I meant no: use

remotes::install_github("r-spatial/stars")

> 
> 
>> Best regards,
>>
>> On 10/23/19 2:22 PM, Andy Teucher wrote:
>>> I’m fairly certain this is a bug in stars; I opened an issue here:
>>> https://github.com/r-spatial/stars/issues/223
>>> <https://github.com/r-spatial/stars/issues/223>
>>>
>>> Cheers,
>>> Andy Teucher
>>>
>>>> On Oct 23, 2019, at 10:42 AM, Andy Teucher <andy.teucher using gmail.com>
>>>> wrote:
>>>>
>>>> Interesting - there’s some rlang/tidy evaluation trickery going on
>>>> there that I couldn’t quite figure out (I think it might be
>>>> searching for yr in the wrong environment), but defining your range
>>>> as a single variable, and putting that in the square brackets seems
>>>> to work for me:
>>>>
>>>> rng <- yr:last_yr
>>>> stars_window = ci_stars[,,,rng]
>>>>
>>>>
>>>>> On Oct 23, 2019, at 10:20 AM, Micha Silver <tsvibar using gmail.com> wrote:
>>>>>
>>>>>
>>>>> On 23/10/2019 19:30, Andy Teucher wrote:
>>>>>> Hi Micha,
>>>>>>
>>>>>> I can see two problems immediately with your code:
>>>>>> 1. you are using a double-colon (yr::last_yr) - the double colon
>>>>>> is used for looking for an object in a package, so it is looking
>>>>>> for object ‘yrs’ in package ‘yr’, which obviously doesn’t make
>>>>>> sense. Use a single colon to create a range (like you did with 2:6)
>>>>>> 2.  the object ‘last_yr’ is never defined, so even if you used a
>>>>>> single colon to define the range yr:last_yr, it would fail as it
>>>>>> would not be able to find object ‘last_yr’
>>>>>>
>>>>> Thanks,
>>>>>
>>>>> I fixed those typos (corrected script attached) and I still get
>>>>> this error:
>>>>>
>>>>>
>>>>> micha using tp480:R$ Rscript stars_window.R
>>>>> Loading required package: abind
>>>>> Loading required package: sf
>>>>> Linking to GEOS 3.7.1, GDAL 2.4.0, PROJ 5.2.0
>>>>> Error in eval(rlang::expr(x[[i]][!!!args])) : object 'yr' not found
>>>>> Calls: RunMK -> [ -> [.stars -> structure -> eval -> eval
>>>>> Execution halted
>>>>>
>>>>>
>>>>>
>>>>>> Cheers,
>>>>>> Andy Teucher
>>>>>>
>>>>>>> On Oct 23, 2019, at 8:28 AM, Micha Silver <tsvibar using gmail.com
>>>>>>> <mailto:tsvibar using gmail.com>> wrote:
>>>>>>>
>>>>>>> I am trying to run a function (mk.test to find MannKendall
>>>>>>> trends) using st_apply over a "rolling" window for a time series
>>>>>>> of rasters in a stars object.
>>>>>>> When I use subscript notation to slice out the window dimension
>>>>>>> with a looping variable I get an error:
>>>>>>>
>>>>>>> Error in loadNamespace(name) : there is no package called ‘yr’
>>>>>>> Calls: [ ... tryCatch -> tryCatchList -> tryCatchOne -> <Anonymous>
>>>>>>> Execution halted
>>>>>>>
>>>>>>> However If I replace the subscript with integers it works fine.
>>>>>>> (see attached)
>>>>>>> What is the correct way to work this out?
>>>>>>>
>>>>>>> Attached is a reprex with a small subset of my data. (The script
>>>>>>> starts with a long structure, code is at the end)
>>>>>>>
>>>>>>> Thanks
>>>>>>> -- 
>>>>>>> Micha Silver
>>>>>>> Ben Gurion Univ.
>>>>>>> Sde Boker, Remote Sensing Lab
>>>>>>> cell: +972-523-665918
>>>>>>> <stars_window.R>_______________________________________________
>>>>>>> R-sig-Geo mailing list
>>>>>>> R-sig-Geo using r-project.org <mailto:R-sig-Geo using r-project.org>
>>>>>>> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>>>>> -- 
>>>>> Micha Silver
>>>>> Ben Gurion Univ.
>>>>> Sde Boker, Remote Sensing Lab
>>>>> cell: +972-523-665918
>>>>>
>>>>> <stars_window.R>
>>>
>>>     [[alternative HTML version deleted]]
>>>
>>> _______________________________________________
>>> R-sig-Geo mailing list
>>> R-sig-Geo using r-project.org
>>> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>>>
>>
>> _______________________________________________
>> R-sig-Geo mailing list
>> R-sig-Geo using r-project.org
>> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
> 

-- 
Edzer Pebesma
Institute for Geoinformatics
Heisenbergstrasse 2, 48151 Muenster, Germany
Phone: +49 251 8333081

-------------- next part --------------
A non-text attachment was scrubbed...
Name: pEpkey.asc
Type: application/pgp-keys
Size: 3110 bytes
Desc: not available
URL: <https://stat.ethz.ch/pipermail/r-sig-geo/attachments/20191026/c91218d0/attachment.bin>