pytroll/mpop

Meteorological Post Processing Package. This package is deprecated, please use satpy instead.

This is the Meteorological Post-Processing Package (MPoP).

A recent build of the documentation is available at http://mpop.readthedocs.org/.

mpop's Issues

DNB support in compact viirs reader

Remove unreferenced blobs/clean up repository

Currently the size of unpacked repository is about 28Mb. A big contributor to the size is the etc/world_map.ascii file, which is removed long time ago. By removing the blobs related to it the size of the (unpacked) repository drops to 8Mb.

Unfortunately removing the blob and all references using git filter-branch:

git filter-branch --index-filter 'git rm --cached --ignore-unmatch etc/world_map.ascii' HEAD

rewrites history and changes commit SHAs, therefore it might be necessary for the developers to start with a clean clone in order to avoid conflicts and pushing dirty history back.

Consider doing cleaning exercise before releasing mpop v2.0.

AttributeError: 'module' object has no attribute 'load'

This is my problem (during compiling example from readthedocs):
Traceback (most recent call last):
File "test.py", line 9, in
global_data.load([0.6, 0.8, 10.8], area_extent=europe.area_extent)
File "/usr/local/lib/python2.7/dist-packages/mpop-v0.13.0-py2.7.egg/mpop/scene.py", line 471, in load
reader_instance = self._set_reader(reader_name)
File "/usr/local/lib/python2.7/dist-packages/mpop-v0.13.0-py2.7.egg/mpop/scene.py", line 358, in _set_reader
loader = getattr(loader, "load")
AttributeError: 'module' object has no attribute 'load'
Smth wrong with the library?

migrate all readers to new architecture

All the readers in mpop need to be migrated.

• aapp1b
• eps1b
• hdf_eos_l1b
• viirs sdr
• compact viirs
• mipp
• hrpt lvl0
  ...

Add possibility to load several versions of the same dataset in a scene

Loading a CTTH or Cloudtype channel for meteosat impossible

Loading a CTTH or Cloudtype channel for meteosat raises an exception about a region not found.
This is due to a bug in MSG not counting the trailing '\0' in the region name string, leading to region names missing their last letter.

Test failure with Pillow 3.4.2

Test failure with Pillow 3.4.2

  ======================================================================
  ERROR: test_pil_image (mpop.tests.test_image.TestEmptyImage)
  Return an empty PIL image.
  ----------------------------------------------------------------------
  Traceback (most recent call last):
    File "/home/lamby/temp/cdt.20161103212915.XxOo8TNjqe.db.python-mpop/python-mpop-1.3.0/mpop/tests/test_image.py", line 184, in test_pil_image
      pilimg = self.img.pil_image()
    File "/home/lamby/temp/cdt.20161103212915.XxOo8TNjqe.db.python-mpop/python-mpop-1.3.0/mpop/imageo/image.py", line 408, in pil_image
      return Pil.new(self.mode, (0, 0))
    File "/usr/lib/python2.7/dist-packages/PIL/Image.py", line 2021, in new
      _check_size(size)
    File "/usr/lib/python2.7/dist-packages/PIL/Image.py", line 2001, in _check_size
      raise ValueError("Width and Height must be > 0")
  ValueError: Width and Height must be > 0
  
  ----------------------------------------------------------------------
  Ran 63 tests in 0.306s
  
  FAILED (errors=1, skipped=1)

See also

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=843079

and

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=843107

Unable to import channels set(['IR_108']) not very helpful

Unable to import channels set(['IR_108']) should be a bit more explicit when it comes to the cause of the inability to load...

getting the version number

print mpop.version
AttributeError: 'module' object has no attribute 'version'

mpop doesn't give the version. Would be nice to have!

New reader for FCI

The test data is available here:

ftp://ftp.eumetsat.int/pub/OPS/out/test-data/FCI_L1C_Format_Familiarisation/

cf conventions "stere" projection in cfscene is buggy

from Ulrich:
I guess there is a small bug in line:
https://github.com/mraspaud/mpop/blob/pre-master/mpop/satout/cfscene.py#L432
"scale_factor_at_projection_origin": eval( proj_dict.get("x_0", "1.0")
as x_0 is already the false northing.

Test failures on mpop 1.2.1

I'm preparing the new debian package for mpop 1.2.1 and I have the following errors when I run the test suite on debian sid:

======================================================================
FAIL: test_project (tests.test_pp_core.TestPPCore)
Test project
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/antonio/debian/git/build-area/python-mpop-1.2.1/mpop/tests/test_pp_core.py", line 99, in test_project
    self.assertEquals(new_scene.area_id, area2)
AssertionError: None != 'oGuyfXoY'

======================================================================
FAIL: test_init (tests.test_projector.TestProjector)
Creation of coverage.
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/usr/lib/python2.7/dist-packages/mock/mock.py", line 1305, in patched
    return func(*args, **keywargs)
  File "/home/antonio/debian/git/build-area/python-mpop-1.2.1/mpop/tests/test_projector.py", line 76, in test_init
    self.assertEquals(utils.parse_area_file.call_count, 2)
AssertionError: 0 != 2

======================================================================
FAIL: test_project_array (tests.test_projector.TestProjector)
Test the project_array function.
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/usr/lib/python2.7/dist-packages/mock/mock.py", line 1305, in patched
    return func(*args, **keywargs)
  File "/home/antonio/debian/git/build-area/python-mpop-1.2.1/mpop/tests/test_projector.py", line 171, in test_project_array
    data, "a", fill_value=None)
  File "/usr/lib/python2.7/dist-packages/mock/mock.py", line 937, in assert_called_with
    six.raise_from(AssertionError(_error_message(cause)), cause)
  File "/usr/lib/python2.7/dist-packages/six.py", line 718, in raise_from
    raise value
AssertionError: Expected call: ImageContainer(array([[-1.50200251],
       [-0.6686467 ],
       [ 1.18901589]]), 'a', fill_value=None)
Actual call: ImageContainer(array([[-1.50200251],
       [-0.6686467 ],
       [ 1.18901589]]), <MagicMock name='mock.utils.parse_area_file().__getitem__()' id='139760093893840'>, fill_value=None)

Also I have an additional error due fact that the trollsift module is missing.
Please note that trollsift is not listed in the package dependencies in setup.py.

Add proj for raw geostationary data when writing to geotiff

from datetime import datetime, timedelta


geosSettings = {
                "met10": {'sat':"meteosat", 'num':"10", 'instr':"seviri", 'ir':'IR_108', 'freq': 15 }
}
sat = "met10"
fOut = "tst.tif"

cycles = 1
dateNow = datetime.now()
t = dateNow - timedelta(minutes=dateNow.minute % 15, seconds=dateNow.second, microseconds=dateNow.microsecond) - timedelta(15*cycles/(24.*60.))

geosSettings[sat]['meta'] = GeostationaryFactory.create_scene(geosSettings[sat]['sat'], 
                                             geosSettings[sat]['num'], 
                                             geosSettings[sat]['instr'], time_slot=t)

geosSettings[sat]['meta'].load([geosSettings[sat]['ir']])
img = geosSettings[sat]['meta'].image.ir108()
img.save(fOut)

Results in no proj information is written to tst.tif. The proj (and proj4 string) is present in meteosat10.cfg, but unless the scene is projected first (with foo.project("projection") ), this information is omitted. On my wishlist to include the raw geos projection string when writing to geotiff even if project has not been called.

Best wishes
Torleif

Metop GDS calibration in eps_l1b reader - inconsistent to DR data

I have seen that there is a difference in the ch3A (1.6 micron) reflectances between the Metop GDS and AAPP direct readout data. I read the Metop GDS with eps_l1b and the DR data with aapp1b.

The GDS reflectances are higher, see figure here

Reading MODIS combined data

Trying to figure out how to do read the MODIS MCD12Q2 using MPOP. Any idea?

Wrong NaN check in satin/helper_functions.py

Hi Martin,
I've experienced some trouble when reading HRIT in conjunction with trollduction and area definitions of different extent.. I think in helper_functions.py the NaN-check in function boundaries_to_extent should be fixed.

for i in range(4):
    if extent[i] is np.nan:
        extent[i] = default_extent[i]

should be replaced with

for i in range(4):
    if np.isnan(extent[i]):
        extent[i] = default_extent[i]

Best regards,
Christian

Remove obsolete sentinel1-feature branch

@loerum going to check if it's actually obsolete.

test_boundaries_to_extent fails with PROJ 5.0.0

As reported in Debian Bug #892462, test_boundaries_to_extent fails since the update of PROJ to version 5.0.0:

======================================================================
FAIL: test_boundaries_to_extent (mpop.tests.test_satin_helpers.TestSatinHelpers)
Test conversion of area boundaries to area extent.
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/usr/lib/python2.7/dist-packages/mpop/tests/test_satin_helpers.py", line 194, in test_boundaries_to_extent
    correct_values_afgh[i], 2)
AssertionError: 0.0 != 3053894.9120028536 within 2 places

CC: @avalentino

Restructure the internal data as 3d arrays

In order to be able to extract profiles for example it would be great if the internal data for the channels could be stored as slices in 3d arrays.

syntax error in sarx.py

syntax error in mpop/instruments/sarx.py

Mask and bracket are missing in line 61:

    data = np.ma.numpy(data, mask=

python 3 compatibility

We need to support python 3 (and 2.6 onwards)