Open jfigui opened 1 year ago
Looking at the code I don't see why it would be ignored. radar.Ai
calls scatter.ext_xsect
which does behave differently (uses a different element of the matrix S
) based on h_pol
. Could you post a working example of that replicates the problem?
diam_min = 0.1 diam_max = 7. step = 0.1 diam = np.arange(diam_min, diam_max+step, step) num_points = diam.size canting_angle = 10.
geom_back = (90.0, 90.0, 0.0, 180.0, 0.0, 0.0) geom_forw = (90.0, 90.0, 0.0, 0.0, 0.0, 0.0)
wavelength = tmatrix_aux.wl_C m = refractive.m_w_20C[wavelength] scatterer = Scatterer(wavelength=wavelength, m=m) scatterer.orient = orientation.orient_averaged_fixed scatterer.or_pdf = orientation.gaussian_pdf(canting_angle) scatterer.psd_integrator = PSDIntegrator() scatterer.psd_integrator.axis_ratio_func = (lambda diam: 1.0/tmatrix_aux.dsr_thurai_2007(diam)) scatterer.psd_integrator.D_max = diam_max scatterer.psd_integrator.num_points = num_points scatterer.psd_integrator.geometries = (geom_back, geom_forw) scatterer.psd_integrator.init_scatter_table(scatterer, angular_integration=True, verbose=True)
psd = GammaPSD(D0=3, Nw=10e3, mu=0, D_max=diam_max)
scatterer.psd = psd
scatterer.set_geometry(geom_forw)
Ah = radar.Ai(scatterer)
Av = radar.Ai(scatterer, False)
@jleinonen,
radar.Ai(scatterer, h_pol=True) calls scatter.ext_xsect(scatterer, h_pol=h_pol)
If the PSD integrator is not None ext_xsect returns:
scatterer.psd_integrator.get_angular_integrated( scatterer.psd, scatterer.get_geometry(), "sca_xsect")
but in this call the information of which polarization you want is lost. By default you get the horizontal polarization regardless of the intended polarization.
Am I right or I am missing something?
Looks like there's indeed a bug. I'm working on a fix.
Thanks a lot!
Hi @jfigui, could you check if it works for you with the latest commit?
Hi @jleinonen ,
It works for me now. See the output:
Could you make a new release soon? I work with the pypi package usually
I don't manage the conda package (someone created it without my knowledge) so I have little control over it. As I understand it gets updated when I create a new release.
I think if you create a new Pypi package you should get a new conda package at some point. At least this is how it works for us using conda-forge
Thanks for being so reactive. Let me know when the new release is ready
Version 0.3.3 is now on PyPI. The conda package should eventually update automatically, as mentioned.
Great!
Thanks a lot for being so reactive. For info I am using pytmatrix in a software package to compute the scattering properties of different hydrometeors. The code is now public and you can find it here: https://github.com/openradar/hydroscatt
I would be grateful if you have a look and report any bug.
I think if you create a new Pypi package you should get a new conda package at some point. At least this is how it works for us using conda-forge
Hi, i tried conda-forge the pytmatrix module using conda
install conda-forge::pytmatrix
But i keep having the same issue : PackagesNotFoundError: The following packages are not available from current channels:
Current channels:
To search for alternate channels that may provide the conda package you're looking for, navigate to
https://anaconda.org
and use the search bar at the top of the page.
I also downloaded the file and conda
install it, the install work, i can see the package in my conda list, but the Channel is
Hi @jleinonen ,
When I compute Ah and Av of rain with a prescribed canting angle using the psd_integrator I get exactly the same results for Ah and for Av. In fact the parameter h_pol in radar.Ai(scatterer, h_pol=False) seems to be ignored when using the integrator.
I am missing something or is this indeed a bug in the code?
Cheers,
@jfigui