Closed sfinkens closed 2 years ago
How did you install Basemap and it's accompanying data? The data was updated in 1.1.0 as well as a few tweaks to the loading code. I never tested the loading code against the old data, so there might be some incompatibilities.
On Jul 4, 2017 12:36 PM, "Stephan Finkensieper" notifications@github.com wrote:
Defining a Basemap in geostationary projection,
from mpl_toolkits.basemap import Basemap Basemap(projection='geos', lon_0=0)
fails in basemap-1.1.0 with the following error:
Traceback (most recent call last):
File "
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/matplotlib/basemap/issues/361, or mute the thread https://github.com/notifications/unsubscribe-auth/AARy-E7eYi2cB2pe4U9YHz63KXGKMimXks5sKmoLgaJpZM4ONi57 .
I used
pip install --prefix=$PREFIX git+https://github.com/matplotlib/basemap.git
Do you have the same problem if installing from conda-forge?
On Jul 5, 2017 3:04 AM, "Stephan Finkensieper" notifications@github.com wrote:
I used
pip install --prefix=$PREFIX git+https://github.com/matplotlib/basemap.git
— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/matplotlib/basemap/issues/361#issuecomment-313019253, or mute the thread https://github.com/notifications/unsubscribe-auth/AARy-HCnKU5gCAyBeEVStKmaunkPs2yDks5sKzV6gaJpZM4ONi57 .
Unfortunately I can't use conda due to legal issues...
uhm, ok. I take it that there is a long an complicated story behind that. I'll see if I can reproduce the problem, but right now, conda is the preferred method of installing packages that have non-python dependencies (namely, libgeos).
On Mon, Jul 10, 2017 at 3:03 AM, Stephan Finkensieper < notifications@github.com> wrote:
Unfortunately I can't use conda due to legal issues...
— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/matplotlib/basemap/issues/361#issuecomment-314022571, or mute the thread https://github.com/notifications/unsubscribe-auth/AARy-FMTcoQi3sSrw_8435e_hCL9za5vks5sMczJgaJpZM4ONi57 .
Indeed :) Thank you very much for your effort! I linked basemap against geos-3.6.1, whose test suite passes.
basemap's test suite? yeah, that isn't very definitive because it isn't extensive at all. The better exercise is to run the examples (needs netcdf4 package, though).
On Mon, Jul 10, 2017 at 10:22 AM, Stephan Finkensieper < notifications@github.com> wrote:
Indeed :) Thank you very much for your effort! I linked basemap against geos-3.6.1, whose test suite passes.
— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/matplotlib/basemap/issues/361#issuecomment-314120743, or mute the thread https://github.com/notifications/unsubscribe-auth/AARy-PREUXRx9vqdfhi-xnGjbEjJ6DCyks5sMjOIgaJpZM4ONi57 .
Aha, the geos definition in examples/test.py
,
m = Basemap(projection='geos',
rsphere=(6378137.00,6356752.3142),
resolution='c',
area_thresh=10000.,
lon_0=0,
satellite_height=35785831)
works fine. Removing the rsphere
argument causes the error again.
Interesting. I guess that makes some sense, but perhaps a better default behavior could be defined?
On Tue, Jul 11, 2017 at 3:03 AM, Stephan Finkensieper < notifications@github.com> wrote:
Aha, the geos definition in examples/test.py,
m = Basemap(projection='geos', rsphere=(6378137.00,6356752.3142), resolution='c', area_thresh=10000., lon_0=0, satellite_height=35785831)
works fine. Removing the rsphere argument causes the error again.
— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/matplotlib/basemap/issues/361#issuecomment-314354073, or mute the thread https://github.com/notifications/unsubscribe-auth/AARy-Il9J89UxjaL8-J17vcNphaVdryUks5sMx44gaJpZM4ONi57 .
Hmm, I don't see why a geostationary projection would not be defined for a spherical earth?
it might be expecting a tuple and erroring on a float or something
I'll look into it tomorrow
On Tue, Jul 11, 2017 at 11:24 AM, Stephan Finkensieper < notifications@github.com> wrote:
Hmm, I don't see why a geostationary projection would not be defined for a spherical earth?
— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/matplotlib/basemap/issues/361#issuecomment-314480076, or mute the thread https://github.com/notifications/unsubscribe-auth/AARy-L6VKrCkNSVVOTMhFM8yQjAJtamsks5sM5OrgaJpZM4ONi57 .
just stumbled on this issue myself, any progress?
Same for me: removing the parameter rsphere=(6378137.00,6356752.3142) triggers the error, but including it brings no result. In fact, reducing the satellite's height by a factor of 100 kills my kernel (first time ever!)
I can confirm that this is still the case for Baseline-1.2.0 and geos-3.6.2. Adding rsphere=(6378137.00,6356752.3142) fixes the problem.
Aha, the geos definition in
examples/test.py
,m = Basemap(projection='geos', rsphere=(6378137.00,6356752.3142), resolution='c', area_thresh=10000., lon_0=0, satellite_height=35785831)
works fine. Removing the
rsphere
argument causes the error again.
It worked for me.thank you.
Aha, the geos definition in
examples/test.py
,m = Basemap(projection='geos', rsphere=(6378137.00,6356752.3142), resolution='c', area_thresh=10000., lon_0=0, satellite_height=35785831)
works fine. Removing the
rsphere
argument causes the error again.
It worked for me too. Thank you!
Defining a Basemap in geostationary projection,
fails in basemap-1.1.0 with the following error: