GithubHelp home page GithubHelp logo

gher-uliege / divand.py Goto Github PK

View Code? Open in Web Editor NEW
8.0 8.0 3.0 390 KB

divand performs an n-dimensional variational analysis of arbitrarily located observations (python interface)

License: GNU General Public License v2.0

Python 64.12% Shell 35.88%

divand.py's People

Contributors

alexander-barth avatar ctroupin avatar gher-diva avatar

Stargazers

avatar avatar avatar avatar avatar avatar avatar avatar

Watchers

avatar avatar avatar avatar avatar

Forkers

arnaldorusso ecutolo

divand.py's Issues

Julia Domain Error linked to len parameters

Good morning,

I am trying to interpolate oceanographic data (nutrients concentrations) with DIVAnd through Python using PyJulia but encounter this error :

RuntimeError: <PyCall.jlwrap (in a Julia function called from Python)
JULIA: DomainError with -0.0007120454096863318:
sqrt will only return a complex result if called with a complex argument. Try sqrt(Complex(x)).
Stacktrace:
  [1] throw_complex_domainerror(f::Symbol, x::Float64)
    @ Base.Math .\math.jl:33
  [2] sqrt
    @ .\math.jl:567 [inlined]
  [3] _broadcast_getindex_evalf
    @ .\broadcast.jl:670 [inlined]
  [4] _broadcast_getindex
    @ .\broadcast.jl:643 [inlined]
  [5] _getindex
    @ .\broadcast.jl:667 [inlined]
  [6] _getindex
    @ .\broadcast.jl:666 [inlined]
  [7] _broadcast_getindex
    @ .\broadcast.jl:642 [inlined]
  [8] getindex
    @ .\broadcast.jl:597 [inlined]
  [9] macro expansion
    @ .\broadcast.jl:961 [inlined]
 [10] macro expansion
    @ .\simdloop.jl:77 [inlined]
 [11] copyto!
    @ .\broadcast.jl:960 [inlined]
 [12] copyto!
    @ .\broadcast.jl:913 [inlined]
 [13] copy
    @ .\broadcast.jl:885 [inlined]
 [14] materialize(bc::Base.Broadcast.Broadcasted{Base.Broadcast.DefaultArrayStyle{2}, Nothing, typeof(/), Tuple{Int64, Base.Broadcast.Broadcasted{Base.Broadcast.DefaultArrayStyle{2}, Nothing, typeof(sqrt), Tuple{Matrix{Float64}}}}})
    @ Base.Broadcast .\broadcast.jl:860
 [15] DIVAnd_background(operatortype::Type, mask::BitMatrix, pmn::Tuple{Matrix{Float64}, Matrix{Float64}}, Labs::Tuple{Matrix{Float64}, Matrix{Float64}}, alpha::Vector{Any}, moddim::Vector{Any}, scale_len::Bool, mapindex::Vector{Any}; btrunc::Vector{Any}, coeff_laplacian::Vector{Float64}, coeff_derivative2::Vector{Float64}, mean_Labs::Nothing)
    @ DIVAnd C:\Users\Léo\.julia\packages\DIVAnd\Iwjb1\src\DIVAnd_background.jl:123
 [16] DIVAndrun(operatortype::Type, mask::BitMatrix, pmnin::Tuple{Matrix{Float64}, Matrix{Float64}}, xiin::Tuple{Matrix{Float64}, Matrix{Float64}}, x::Tuple{Vector{Float64}, Vector{Float64}}, f::Vector{Float64}, lin::Tuple{Float64, Float64}, epsilon2::Int64; velocity::Tuple{}, primal::Bool, factorize::Bool, tol::Float64, maxit::Int64, minit::Int64, constraints::Tuple{}, inversion::Symbol, moddim::Vector{Any}, fracindex::Matrix{Float64}, alpha::Vector{Any}, keepLanczosVectors::Int64, compPC::typeof(DIVAnd.DIVAnd_pc_none), progress::DIVAnd.var"#285#287", fi0::Matrix{Float64}, f0::Vector{Float64}, alphabc::Float64, scale_len::Bool, btrunc::Vector{Any}, MEMTOFIT::Float64, topographyforfluxes::Tuple{}, fluxes::Tuple{}, epsfluxes::Int64, epsilon2forfractions::Int64, RTIMESONESCALES::Tuple{}, QCMETHOD::Tuple{}, coeff_laplacian::Vector{Float64}, coeff_derivative2::Vector{Float64}, mean_Labs::Nothing)
    @ DIVAnd C:\Users\Léo\.julia\packages\DIVAnd\Iwjb1\src\DIVAndrun.jl:53
 [17] DIVAndrun(operatortype::Type, mask::BitMatrix, pmnin::Tuple{Matrix{Float64}, Matrix{Float64}}, xiin::Tuple{Matrix{Float64}, Matrix{Float64}}, x::Tuple{Vector{Float64}, Vector{Float64}}, f::Vector{Float64}, lin::Tuple{Float64, Float64}, epsilon2::Int64)
    @ DIVAnd C:\Users\Léo\.julia\packages\DIVAnd\Iwjb1\src\DIVAndrun.jl:45
 [18] #DIVAndrun#290
    @ C:\Users\Léo\.julia\packages\DIVAnd\Iwjb1\src\DIVAndrun.jl:298 [inlined]
 [19] DIVAndrun
    @ C:\Users\Léo\.julia\packages\DIVAnd\Iwjb1\src\DIVAndrun.jl:298 [inlined]
 [20] #DIVAndrun#289
    @ C:\Users\Léo\.julia\packages\DIVAnd\Iwjb1\src\DIVAndrun.jl:152 [inlined]
 [21] DIVAndrun(::Matrix{Bool}, ::Tuple{Matrix{Float64}, Matrix{Float64}}, ::Tuple{Matrix{Float64}, Matrix{Float64}}, ::Tuple{Vector{Float64}, Vector{Float64}}, ::Vector{Float64}, ::Tuple{Float64, Float64}, ::Int64)
    @ DIVAnd C:\Users\Léo\.julia\packages\DIVAnd\Iwjb1\src\DIVAndrun.jl:152
 [22] DIVAndrunfi(::Matrix{Bool}, ::Vararg{Any})
    @ DIVAnd C:\Users\Léo\.julia\packages\DIVAnd\Iwjb1\src\DIVAndrun.jl:157
 [23] invokelatest(::Any, ::Any, ::Vararg{Any}; kwargs::Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}})
    @ Base .\essentials.jl:716
 [24] invokelatest(::Any, ::Any, ::Vararg{Any})
    @ Base .\essentials.jl:714
 [25] _pyjlwrap_call(f::Function, args_::Ptr{PyCall.PyObject_struct}, kw_::Ptr{PyCall.PyObject_struct})
    @ PyCall C:\Users\Léo\.julia\packages\PyCall\7a7w0\src\callback.jl:28
 [26] pyjlwrap_call(self_::Ptr{PyCall.PyObject_struct}, args_::Ptr{PyCall.PyObject_struct}, kw_::Ptr{PyCall.PyObject_struct})
    @ PyCall C:\Users\Léo\.julia\packages\PyCall\7a7w0\src\callback.jl:44>

The parameters used are defined as follow :

lenx = 1/12 * 110000 * np.cos(np.pi * 40 / 180)  # we estimate measures are not drastically different whithin a 1/12° square
leny = 1/12 * 110000
epsilon2 = 1.
mask = np.full(XX.shape, True, dtype=bool)
pm = np.ones(XX.shape) / (XX[0, 1] - XX[0, 0])
pn = np.ones(XX.shape) / (YY[1, 0] - YY[0, 0])
interp_values = DIVAnd(mask, (pm, pn), (XX, YY), (points[:, 0], points[:, 1]), values, (lenx, leny), epsilon2)
print("va", interp_values)

Where XX and YY are the meshgrid to which we interpolate, points are lon and lat measures coordinates and values the measures, all positive.

I observed that the value explained in Domain error decreases when (lenx, leny) increases, but never becomes positive.

Any solution or idea about this ?

Great day,
Léo

All zero results with 3D fields

Hi,
I'm trying to use DIVAnd.py with a 3d field so I just adapted one of the example written in JULIA to check if it works but I just obtain a zero fields. Here is the code I used:

from julia import DIVAnd
import numpy as np

x = np.random.rand(75);
y = np.random.rand(75);
z = np.random.rand(75);
f = np.sin(x*6) * np.cos(y*6)+np.sin(z*6) * np.cos(x*6) ;

testsize=300
testsizez=3
xi,yi,zi = np.meshgrid(np.arange(0,1,1/testsize),np.arange(0,1,1/testsize),np.arange(0,1,1/testsizez));

fref = np.sin(xi*6) * np.cos(yi*6)+np.sin(zi*6) * np.cos(xi*6) ;

mask = np.ones(xi.shape)==1;

pm = np.ones(xi.shape) / (xi[1,2,1]-xi[1,1,1])
pn = np.ones(xi.shape) / (yi[2,1,1]-yi[1,1,1])
po = np.ones(xi.shape) / (zi[1,1,2]-zi[1,1,1])

lenx = 0.5;

epsilon2 = 1.;

va = DIVAnd.DIVAndrunfi(mask,(pm,pn,po),(xi,yi,zi),(x,y,z),f,lenx,epsilon2)

I have also tryed to transpose some of the fields (as in DIVAnd.py) but I still obtain a zero output

va = DIVAnd.DIVAndrunfi(np.transpose(mask),(np.transpose(pm),np.transpose(pn),np.transpose(po)),
(np.transpose(xi),np.transpose(yi),np.transpose(zi)),
(x,y,z),f,lenx,epsilon2)

I am supposing there is a different way in which JULIA and python order the axes, I will do other tests swapping them.

ImportError

Hi,

I followed all steps to install divand.py (on a workstation with Win 8.1 64 bit)
After I downloaded divand.py, I set a system variable:
PYTHONPATH=C:\Python27;C:\Python27\Lib;C:\Python27\DLLs;C:\divand.py\divand\divand.py

I run: python divand_argo.py
I got : ImportError: No module named divand

Cheers!
George

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.