korte issueshttps://git.gfz-potsdam.de/groups/sec23/korte/-/issues2023-09-12T12:23:44+02:00https://git.gfz-potsdam.de/sec23/korte/paleokalmag/-/issues/10Diag is calculated multiple times in corefieldmodel2023-09-12T12:23:44+02:00Maximilian SchannerDiag is calculated multiple times in corefieldmodelWe should add diag as an attribute instead of calculating it during dt_BB cov again. This will make inheritance and custom models easier.We should add diag as an attribute instead of calculating it during dt_BB cov again. This will make inheritance and custom models easier.https://git.gfz-potsdam.de/sec23/korte/paleokalmag/-/issues/9rename co-lat to colat2023-09-07T13:56:59+02:00Maximilian Schannerrename co-lat to colathttps://git.gfz-potsdam.de/sec23/korte/paleokalmag/-/issues/8save and load is unnecessarily complicated2023-09-12T12:23:30+02:00Maximilian Schannersave and load is unnecessarily complicatedI think it would be much simpler to just pickle the model instead of selecting the attributes and saving stuff as a .npz file...I think it would be much simpler to just pickle the model instead of selecting the attributes and saving stuff as a .npz file...https://git.gfz-potsdam.de/sec23/korte/paleokalmag/-/issues/7read_data docstring is erroneous2023-09-07T13:59:27+02:00Maximilian Schannerread_data docstring is erroneoushttps://git.gfz-potsdam.de/sec23/korte/paleokalmag/-/blob/master/paleokalmag/data_handling.py#L52 states that it's still possible to provide a DataFrame, which is not the case anymorehttps://git.gfz-potsdam.de/sec23/korte/paleokalmag/-/blob/master/paleokalmag/data_handling.py#L52 states that it's still possible to provide a DataFrame, which is not the case anymorehttps://git.gfz-potsdam.de/sec23/korte/paleokalmag/-/issues/6It would be nice to allow sampling from the prior2022-10-11T16:05:07+02:00Maximilian SchannerIt would be nice to allow sampling from the priorImplementation should be straight-forward as a member function of KalmanImplementation should be straight-forward as a member function of KalmanMaximilian SchannerMaximilian Schannerhttps://git.gfz-potsdam.de/sec23/korte/paleokalmag/-/issues/5Add len to Data routine2022-02-03T15:14:25+01:00Maximilian SchannerAdd len to Data routineIt would be nice if `len(Data)` would return the number of records.It would be nice if `len(Data)` would return the number of records.Maximilian SchannerMaximilian Schannerhttps://git.gfz-potsdam.de/sec23/korte/paleokalmag/-/issues/4Docstring of CoreFieldModel.reject_outliers is wrong2022-02-03T15:14:24+01:00Maximilian SchannerDocstring of CoreFieldModel.reject_outliers is wrongMaximilian SchannerMaximilian Schannerhttps://git.gfz-potsdam.de/sec23/korte/paleokalmag/-/issues/3Printing of ChunkedData gives error for empty intervals2022-02-03T15:14:24+01:00Maximilian SchannerPrinting of ChunkedData gives error for empty intervalsThis is a bug in Data's __repr__ routine, that fails if Data doesn't contain any actual data...This is a bug in Data's __repr__ routine, that fails if Data doesn't contain any actual data...Maximilian SchannerMaximilian Schannerhttps://git.gfz-potsdam.de/sec23/korte/pymagglobal/-/issues/22We return the sqrt of the covariance matrix, which is nonsense...2021-11-19T15:12:15+01:00Maximilian SchannerWe return the sqrt of the covariance matrix, which is nonsense...Check the code at [core.py](https://git.gfz-potsdam.de/sec23/korte/pymagglobal/-/blob/master/pymagglobal/core.py#L603). We may have to return the diagonal and reshape, to give the same return as for 'dif'. Maybe this has consequences els...Check the code at [core.py](https://git.gfz-potsdam.de/sec23/korte/pymagglobal/-/blob/master/pymagglobal/core.py#L603). We may have to return the diagonal and reshape, to give the same return as for 'dif'. Maybe this has consequences elsewhere, so better check thoroughly...Maximilian SchannerMaximilian Schannerhttps://git.gfz-potsdam.de/sec23/korte/paleokalmag/-/issues/2Data allocation should not generate diagonal array.2021-11-01T10:35:25+01:00Maximilian SchannerData allocation should not generate diagonal array.When setting up the data class, the error matrix is already associated to the Data object as a full 2d array, although we only ever use diagonal ones. See [here](https://git.gfz-potsdam.de/sec23/korte/paleokalmag/-/blob/master/paleokalma...When setting up the data class, the error matrix is already associated to the Data object as a full 2d array, although we only ever use diagonal ones. See [here](https://git.gfz-potsdam.de/sec23/korte/paleokalmag/-/blob/master/paleokalmag/data_handling.py#L220). Only this line prohibits reading large data files as a Data object. We should change this, but have to modify the inversion, where the errors are accessed...Maximilian SchannerMaximilian Schannerhttps://git.gfz-potsdam.de/sec23/korte/pymagglobal/-/issues/21Documentation for equi_sph is wrong2021-10-20T15:47:23+02:00Clara FischerDocumentation for equi_sph is wrongAzimutal and polar angle are exchanged...Azimutal and polar angle are exchanged...https://git.gfz-potsdam.de/sec23/korte/paleokalmag/-/issues/1Imported Data contains NaN2021-10-20T10:34:49+02:00Clara FischerImported Data contains NaNIf the imported dataset contains NaNs in a column named 'dt', they will be kept in the output. See [here](https://git.gfz-potsdam.de/sec23/korte/paleokalmag/-/blob/master/paleokalmag/data_handling.py#L223).If the imported dataset contains NaNs in a column named 'dt', they will be kept in the output. See [here](https://git.gfz-potsdam.de/sec23/korte/paleokalmag/-/blob/master/paleokalmag/data_handling.py#L223).Maximilian SchannerMaximilian Schannerhttps://git.gfz-potsdam.de/sec23/korte/pymagglobal/-/issues/20Additional models2021-08-25T13:46:01+02:00Maximilian SchannerAdditional modelsAdditional models were distributed via the gfz ftp server, but many browsers stopped supporting ftp natively. We should think of an alternative way of distributing these models, maybe a second repo?Additional models were distributed via the gfz ftp server, but many browsers stopped supporting ftp natively. We should think of an alternative way of distributing these models, maybe a second repo?https://git.gfz-potsdam.de/sec23/korte/pymagglobal/-/issues/19Upload should take place after testing!2021-07-28T10:50:25+02:00Maximilian SchannerUpload should take place after testing!Currently the package is built, uploaded and then tested. This is very bad. We should build it and pass the artifact to a testing stage which passes the tested built to an uploading job...Currently the package is built, uploaded and then tested. This is very bad. We should build it and pass the artifact to a testing stage which passes the tested built to an uploading job...https://git.gfz-potsdam.de/sec23/korte/pymagglobal/-/issues/18The EOU messages look ugly in the command line2020-07-29T15:52:17+02:00Maximilian SchannerThe EOU messages look ugly in the command lineI get that it's nice for the documentation, but maybe there is a way to omit the rst in the actual python code... seeing
```console
Example of use: :command:`pymagglobal dipole CALS10k.2` .. image::
./pic_dip.png
```
in the command line ...I get that it's nice for the documentation, but maybe there is a way to omit the rst in the actual python code... seeing
```console
Example of use: :command:`pymagglobal dipole CALS10k.2` .. image::
./pic_dip.png
```
in the command line help message is ugly...Stefan MauerbergerStefan Mauerbergerhttps://git.gfz-potsdam.de/sec23/korte/pymagglobal/-/issues/17New tests should actually pass, not fail2020-07-24T15:00:29+02:00Maximilian SchannerNew tests should actually pass, not failShouldn't [these tests](https://gitext.gfz-potsdam.de/arthus/pymagglobal/-/blob/master/pymagglobal/tests/test_argument_parser.py#L85) actually pass, since [here](https://gitext.gfz-potsdam.de/arthus/pymagglobal/-/blob/master/pymagglobal/...Shouldn't [these tests](https://gitext.gfz-potsdam.de/arthus/pymagglobal/-/blob/master/pymagglobal/tests/test_argument_parser.py#L85) actually pass, since [here](https://gitext.gfz-potsdam.de/arthus/pymagglobal/-/blob/master/pymagglobal/__main__.py#L125) we set the attrs anyway???Stefan MauerbergerStefan Mauerbergerhttps://git.gfz-potsdam.de/sec23/korte/pymagglobal/-/issues/16Sequence of components2020-07-24T11:24:08+02:00Stefan MauerbergerSequence of componentsThe argument `--type` could be renamed to `--comps` or `--components` and accept a string.
The string should then consist of any combination of `DIHFNEZ`. The horizontal component is actually only of interest for observatory data.The argument `--type` could be renamed to `--comps` or `--components` and accept a string.
The string should then consist of any combination of `DIHFNEZ`. The horizontal component is actually only of interest for observatory data.https://git.gfz-potsdam.de/sec23/korte/pymagglobal/-/issues/15Cut-off degree2020-07-24T11:24:05+02:00Stefan MauerbergerCut-off degreeIn some cases it might be useful to have an optional argument `--lmax` e.g for plotting the spatial power spectrum or coefficients.In some cases it might be useful to have an optional argument `--lmax` e.g for plotting the spatial power spectrum or coefficients.https://git.gfz-potsdam.de/sec23/korte/pymagglobal/-/issues/14Temporal Power Spectrum2020-07-24T11:24:01+02:00Stefan MauerbergerTemporal Power SpectrumFor a spline base, there should be an explicit form for the spectrum.For a spline base, there should be an explicit form for the spectrum.https://git.gfz-potsdam.de/sec23/korte/pymagglobal/-/issues/13Defaut discretization for time-series2020-07-29T15:58:38+02:00Stefan MauerbergerDefaut discretization for time-seriesWouldn't it be better if we used the konts of the coefficients as default discretization?Wouldn't it be better if we used the konts of the coefficients as default discretization?