pyart
pyart copied to clipboard
ARM-DOE
Reading a file file gridded from Radx
I have used radx to grid some cfradial files. I would now like to read them in with grid_io.
I had a go at adding a few things into translate between the variable names of the two file types and was successful at a gradx gridded file in but I am still unable to plot the data using display.plot_grid.
I believe the issues is in the setting up of the origin_* variables and projection dict as the code crases when I try display.get_basemap().
Attached #is the file I am trying to open and the ipython notebook I am using to explore this issue. notebookandtestfile.zip
@rrniii I'm not super familiar with the Grid side, but from the error message and looking at the reference manual I think you need to create a basemap instance and add it to the GridMapDisplay after you create it. The following added to your example code creates a basemap, but I still don't get any output on the plot.
corners = [dset.variables['lon0'][:].min(), dset.variables['lon0'][:].max(),
dset.variables['lat0'][:].min(), dset.variables['lat0'][:].max()]
lon0 = corners[0] + np.abs(corners[1]-corners[0])/2
lat0 = corners[2] + np.abs(corners[3]-corners[2])/2
fig, ax = plt.subplots(1, 1, figsize=(15, 8))
bm = display.plot_basemap(lat_lines=np.linspace(corners[0], corners[1], 5),
lon_lines=np.linspace(corners[2], corners[3], 5), resolution='l',
auto_range=False, min_lon=corners[0], max_lon=corners[1],
min_lat=corners[2], max_lat=corners[3],
projection='tmerc', lat_0=lat0, lon_0=lon0, ax=ax)
p = display.plot_grid('dBZ', level=20, ax=ax)
Thanks @nguy. Your suggestion bumped me into the correct direction. I now have a solution that works. It all came to making my own projParams dictionary from the info in the radx file and getting the correct units. To plot, I still have to set up a basemap (i.e., it does not plot in one step like a grid made with pyart's internal functions. I think this possibly is due to arrays versus lists versus tuples or and possibly some missing info in the proj Params).
I attach an example file from before, a python notebook where I work through all the issues and an altered version of grid_io that was the additional function defined that reads in this file and allows for expansion to further projections.
readingradxgriddeddatawithpyart_10Aug2016.zip
Glad it's working @rrniii !
@jjhelmus Is it feasible to provide either a conversion utility for RadX to Py-ART? Or does this exist, the documentation for the radx portion isn't very clear. A related thing I am encountering is to figure out a way to pipe Sigmet files through Radx to apply geolocation corrections for mobile platforms. I can do this in a separate step, and this may be the way to go, but wondered about making it more seamless.
@rrniii I just looked at the examples you provided. How general is the radx_grid routine you used? You might consider submitting this as a pull request so that it becomes part of the base code. Good candidate for the pyart.aux_io methods?
@rrniii Really nice work! It looks like Py-ART should be capable of reading in grid files created with RadX. I tracked down an issue in Py-ART's coordinate transform code, #568, which I think was causing the issue with plotting these grids without providing explicit basemap parameters. With that fix and a few modification to the code you provided, plotting RadX grids using Py-ART can now be done with:
read_radx_grid('./RAINS_uncal_BNG_5km20160607_222951.nc')
display = pyart.graph.GridMapDisplay(grid)
display.plot_basemap()
display.plot_grid('CCOR', level=10)
plt.show()
I have modified your code a bit and would be happy to add this functionality to Py-ART if you are interested. If you would rather submit the change yourself I'd be happy to help you in this process also.
Is it feasible to provide either a conversion utility for RadX to Py-ART? Or does this exist, the documentation for the radx portion isn't very clear. A related thing I am encountering is to figure out a way to pipe Sigmet files through Radx to apply geolocation corrections for mobile platforms. I can do this in a separate step, and this may be the way to go, but wondered about making it more seamless.
This is the first for supporting Radx grid files in Py-ART. Radar files written by Radx should be CfRadial compliant and can be read with pyart.io.read_cfradial. In addition, the pyart.aux_io.read_radx function can be used to pipe Sigmet/Dorade/etc files through Radx and read the result without needed to drop to the command line. The source of that function might help you create a function to apply the geolocations corrections using Radx without using a separate step.
I'll look at how the input command is structured and the possibility of extending the input if needed. Thanks @jjhelmus
@rrniii Wanted to check on this issue. I wrote up a function which can read in NetCDF grid files from RadX based on your code. If you approve I will add it to Py-ART.
Where are we at on this? I know Brenda Dolan from CSU was interested in this.. @rrniii Did this work? @jjhelmus Did this go in a Pull request some where?
A ping on this topic, it was raised again at the AMS Open software short course.
Sorry for being lame.
Ill get on this. Glad there is interest.
On Aug 28, 2017, at 9:00 AM, Nick Guy [email protected] wrote:
A ping on this topic, it was raised again at the AMS Open software short course.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ARM-DOE/pyart/issues/566#issuecomment-325361518, or mute the thread https://github.com/notifications/unsubscribe-auth/APbprjceYBJ0Rroxe8k6Bo013pbKgItFks5scsgAgaJpZM4JgPGL.
@rrniii Is this working for you? If I don't hear back I will close this issue in a few days (clean out time)
Hey Scott, I thought I had this working. I am away on a filed campaign and crunched for time but I think I can get it up there when back.
On 22 May 2018, at 14:33, Scott [email protected] wrote:
@rrniii https://github.com/rrniii Is this working for you? If I don't hear back I will close this issue in a few days (clean out time)
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ARM-DOE/pyart/issues/566#issuecomment-390991511, or mute the thread https://github.com/notifications/unsubscribe-auth/APbprrTFegc-GOR07F-y3tKBlgonzhfJks5t1BOmgaJpZM4JgPGL.
