Hi Raphaël, Yes, you should upgrade to v2.2. This is what we consider as the present stable version. This means that if your problem will show up to be a bug, a bug fix will go into 2.2, but not v2.0.
This also means that you need to make some smaller changes in your matlab files. The changes in ARTS are documented in CHANGES, but unfortunately there is no such file. Changes are only documented in atmlab/ChangeLog. But maybe somebody on the qpack mailing list can summarise the changes needed in Qpack/Qarts to move from v2.0 v2.2? Bye, Patrick On 11/11/15 09:22, [email protected] wrote: > Hi again, > > If I use the path of ARTS given in the error, I get: > > --------------------------------------------------------------- > $ XXX/ARTS/arts-2.0.40/build/src/arts -v > > arts-2.0.40 (Wed May 9 17:41:12 2012) > Compiler: GNU C++ (/usr/bin/c++) > Compile flags: -ftemplate-depth-60 -fopenmp -O2 -g > Features in this build: > Numeric precision: double > OpenMP support: enabled > Documentation server: enabled > Zipped XML support: enabled > NetCDF support: disabled > Disort algorithm: disabled > Include search paths: > XXX/ARTS/arts-2.0.40/includes > --------------------------------------------------------------- > > Bye, > Raphaël > > -----Original Message----- > From: Patrick Eriksson [mailto:[email protected]] > Sent: Dienstag, 10. November 2015 19:52 > To: Monnard Raphaël > Subject: Re: [arts-users] specific ANTENNA_RESPONSE as function of backend > frequencies > > Hej, > > What ARTS version are you using? > > To check: arts -v > > Bye, > > Patrick > > On 11/10/15 15:22, [email protected] wrote: >> Hi again, >> >> As attached, the asked files. Thanks. >> >> Bye, >> Raphaël >> >> -----Original Message----- >> From: Patrick Eriksson [mailto:[email protected]] >> Sent: Dienstag, 10. November 2015 14:29 >> To: Monnard Raphaël; [email protected] >> Subject: Re: [arts-users] specific ANTENNA_RESPONSE as function of >> backend frequencies >> >> Hi again, >> >> I could not see anything obviously wrong in your antenna definition file. >> Seems that I must test run this. >> >> To prepare a test case for me: >> >> In Matlab >> >> >> atmlab('DEBUG',true) >> >> Run your your case, and check the error message for the name of the >> temporary folder created. It should be /tmp/atmlab-XXX >> >> Make a tar of the temporary folder and send it to me. >> >> To re-set things. Remove all /tmp/atmlab-XXX and >> >> >> atmlab('DEBUG',false) >> >> >> This should normally suffice, but please check in cfile.arts that you don't >> include files not found locally in the temporary folder. If you do, please >> add those files (if not e.g. the complete HITRAN catalogue). >> >> Bye, >> >> Patrick >> >> >> On 11/10/15 14:12, [email protected] wrote: >>> Hi Patrick, >>> >>> Thanks for your reply. If ANTENNA_DO is set false, the simulation is >>> working. Therefore, I strongly suppose that this comes from the antenna >>> pattern. >>> >>> Please find the .mat structure who define the antenna response. >>> >>> Bye, >>> Raphaël >>> >>> -----Original Message----- >>> From: Patrick Eriksson [mailto:[email protected]] >>> Sent: Dienstag, 10. November 2015 13:36 >>> To: Monnard Raphaël; [email protected] >>> Subject: Re: [arts-users] specific ANTENNA_RESPONSE as function of >>> backend frequencies >>> >>> Hi Raphaël, >>> >>> You have managed to find a bug, or a loophole in checking of input. A user >>> should never see an assertion. >>> >>> Are you sure that this comes from the antenna pattern? If yes, send me the >>> structure defining the antenna pattern. Either as .mat or .xml. >>> >>> Bye, >>> >>> Patrick >>> >>> >>> On 11/10/15 09:20, [email protected] wrote: >>>> Hi Patrick, >>>> >>>> Thanks for your reply. If I cover the complete range of f_grid, I get the >>>> following error: >>>> >>>> ------------------------------------ >>>> arts: XXX/ARTS/arts-2.0.40/src/interpolation.cc:176: void >>>> gridpos(ArrayOfGridPos&, ConstVectorView, ConstVectorView, const >>>> Numeric&): Assertion `is_size(gp,n_new)' failed. >>>> >>>> XXX/ARTS/arts-2.0.40/build/src/arts -r00 /tmp/atmlab-XXX/cfile.arts: >>>> Aborted >>>> ------------------------------------ >>>> >>>> Do you have any inputs or suggestions? >>>> >>>> Bye, >>>> Raphaël >>>> >>>> -----Original Message----- >>>> From: Patrick Eriksson [mailto:[email protected]] >>>> Sent: Montag, 9. November 2015 21:37 >>>> To: Monnard Raphaël; [email protected] >>>> Subject: Re: [arts-users] specific ANTENNA_RESPONSE as function of >>>> backend frequencies >>>> >>>> Hi Raphaël, >>>> >>>> You simply need to define the antenna pattern over a wider frequency >>>> range. According to the error message you need to add extend the frequency >>>> coverage with 1.25 and 2 GHz in lower/upper end. Note that it is not >>>> sufficient to cover the centre frequency of the channels. Details change >>>> with the set-up, but I guess that you need to cover the complete range of >>>> f_grid. >>>> >>>> Bye, >>>> >>>> Patrick >>>> >>>> >>>> >>>> On 11/09/15 13:01, [email protected] wrote: >>>>> Hi, >>>>> >>>>> Thanks for your reply. I have tried to implement, with the >>>>> structure defined as attached file, and I obtain the following error: >>>>> >>>>> Run-time error in method: sensor_responseAntenna The frequency grid >>>>> of *antenna_response is too narrow. It must cover all considered >>>>> frequencies (*f_grid*), if the length is > 1. The grid needs to be >>>>> expanded with 1.24e+09 Hz in the lower end. The frequency grid of >>>>> *antenna_response is too narrow. It must cover all considered >>>>> frequencies (*f_grid*), if the length is > 1. The grid needs to be >>>>> expanded with 2e+09 Hz in the upper end. >>>>> >>>>> Do you have any inputs? >>>>> >>>>> Thank you for your reply, Raphaël >>>>> >>>>> -----Original Message----- From: Patrick Eriksson >>>>> [mailto:[email protected]] Sent: Donnerstag, 5. November >>>>> 2015 16:57 To: Monnard Raphaël; [email protected] Subject: >>>>> Re: [arts-users] specific ANTENNA_RESPONSE as function of backend >>>>> frequencies >>>>> >>>>> Hi Raphael, >>>>> >>>>> The answer is yes? One of the dimensions in "antenna response" is >>>>> frequency! The format is described by e.g. doing: >>>>> >>>>> arts -d antenna_response >>>>> >>>>> The output is found below. You just basically include the antenna >>>>> pattern for the frequencies you have the antenna pattern defined in >>>>> the data structure. >>>>> >>>>> Please not that the frequencies in backend channel response is >>>>> somewhat different, they are relative frequencies with respect to >>>>> the channel centre position(s). And this one reason why this is an >>>>> array (and not a GriddedField2). >>>>> >>>>> Bye, >>>>> >>>>> Patrick >>>>> >>>>> >>>>> >>>>> *------------------------------------------------------------------ >>>>> - >>>>> * >>>>> >>>>> >>>> Workspace variable = antenna_response >>>>> ------------------------------------------------------------------- >>>>> - >>>>> - >>>>> >>>>> The antenna pattern/response. >>>>> >>>>> This WSV describes the antenna response as a function of >>>>> polarisation (pol), frequencue (f), zenith angle (za) and azimuth angle >>>>> (aa). >>>>> >>>>> Polarisation dimension: If this dimension has size 1, the data are >>>>> applied for all polarisations of concern. The data are otherwise >>>>> used in sequential order. This signifies that, in general, the >>>>> first polarisation "layer" corresponds to the first stokes dimension etc. >>>>> An exception is if a polarisation rotation has been applied. In any >>>>> case, it is up to the user to ensure that polarisations are >>>>> consistently defined. >>>>> >>>>> Frequency dimension: If this dimension has size 1, the data are >>>>> applied for all frequencies of concern. The given frequency must be >>>>> inside the frequency range of concern. A linear interpolation is >>>>> otherwise applied. >>>>> >>>>> Zenith angle dimension: This dimension must always have a size >= 2 >>>>> The response outside covered grid range is treated as zero. If >>>>> *antenna_dim* equals 1, the data should correspond to the response >>>>> integrated in the azimuthal direction. >>>>> >>>>> Azimuth angle dimension: If *antenna_dim* equals 1, this dimension >>>>> must have size 1. A size >= 2 is otherwise required. The response >>>>> outside covered grid range is treated as zero. >>>>> >>>>> Usage: Set by the user. >>>>> >>>>> Dimensions: GriddedField4: ArrayOfString field_names[N_pol] Vector >>>>> f_grid[N_f] Vector za_grid[N_za] Vector aa_grid[N_aa] Tensor4 >>>>> data[N_pol][N_f][N_za][N_aa] >>>>> >>>>> >>>>> ------------------------------------------------------------------- >>>>> - >>>>> - >>>>> >>>>> >>>> Group = GriddedField4 >>>>> *------------------------------------------------------------------ >>>>> - >>>>> * >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> On 11/05/15 16:38, [email protected] wrote: >>>>>> Hi, >>>>>> >>>>>> I try to simulate a sensor with a specific ANTENNA_RESPONSE as >>>>>> function of backend frequencies. My goal is to perform only one >>>>>> simulation for all backend frequencies. Unfortunately, the format >>>>>> of ANTENNA_RESPONSE is 'GriddedField4', and not >>>>>> '_ArrayOf_GriddedField1' as for BACKEND_CHANNEL_RESPONSE. >>>>>> >>>>>> Therefore, I obtain the following error when I use the H structure >>>>>> as defined in the attached file: >>>>>> >>>>>> Attempt to reference field of non-structure array. >>>>>> >>>>>> Error in xmlWriteGFieldWrapper (line 27) >>>>>> >>>>>> attrlist = xmlAddAttribute ([], 'name', data.gridnames{i}); >>>>>> >>>>>> Is it possible to overcome a such issue? >>>>>> >>>>>> Thank you for your reply. >>>>>> >>>>>> >>>>>> >>>>>> _______________________________________________ arts-users mailing >>>>>> list [email protected] >>>>>> https://www.sat.ltu.se/mailman/listinfo/arts-users >>>>>> >>>>>> >>>>>> >>>>>> _______________________________________________ arts-users mailing >>>>>> list [email protected] >>>>>> https://www.sat.ltu.se/mailman/listinfo/arts-users >>>> _______________________________________________ >>>> arts-users mailing list >>>> [email protected] >>>> https://www.sat.ltu.se/mailman/listinfo/arts-users >>>> >>>> >>>> >>>> _______________________________________________ >>>> arts-users mailing list >>>> [email protected] >>>> https://www.sat.ltu.se/mailman/listinfo/arts-users >>>> >>>> >>>> _______________________________________________ >>>> arts-users mailing list >>>> [email protected] >>>> https://www.sat.ltu.se/mailman/listinfo/arts-users > _______________________________________________ > arts-users mailing list > [email protected] > https://www.sat.ltu.se/mailman/listinfo/arts-users > _______________________________________________ qpack mailing list [email protected] https://www.sat.ltu.se/mailman/listinfo/qpack
