[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Support #TWK-728245]: Re: Python support post AMS workshop

This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.

  • Subject: [Support #TWK-728245]: Re: Python support post AMS workshop
  • Date: Mon, 22 Mar 2021 16:47:09 -0600
Greetings!

No, MetPy does not rely on the same code as NCL to compute cape. Unfortunately, 
the use of dewpoint or relatively humidity doesn't affect the problem--in fact, 
MetPy's `dewpoint_from_relative_humidity` function works fine on grids of data.

The reason NCL's works quickly is because it is written manually in C/FORTRAN 
and loops over values in the compiled language. In Python, function calls are 
slow, so looping over a 2D array in Python and calling a function (e.g. some 
kind of calculation) will be a slow process. Thus, we have NumPy arrays that 
help speed up many calculations by performing the entire calculation across the 
entire array at once (and NumPy handles looping under the hood in C). 
Unfortunately, this only works for calculations that can readily be expressed 
as array operations. The calculations involved in calculating CAPE (integrals, 
finding per-grid point LCL, LFC, and EL points) don't lend themselves to 
NumPy's looping capabilities, so it requires doing it manually. We are planning 
on addressing this in the future, but there's nothing we have to offer besides 
manually looping at the moment.

Cheers,

Ryan

> Hi Ryan,
> 
> Sorry,  misclicked the send button.
> 
> I noticed that NCL's cape function used relative humidity as input,
> not the dew-point. Should this
> make the array operation a bit more doable with MetPy too?
> 
> >
> >     Hi Ryan,
> >
> >     plt.savefig(...) works.  Thank you!
> >
> >     I've tested that NCL can handle 2-dimensional arrays for cape and
> > helicity calculation, not sure if you may use their core?
> >


Ticket Details
===================
Ticket ID: TWK-728245
Department: Support Python
Priority: Normal
Status: Closed
===================
NOTE: All email exchanges with Unidata User Support are recorded in the Unidata 
inquiry tracking system and then made publicly available through the web.  If 
you do not want to have your interactions made available in this way, you must 
let us know in each email you send to us.