Edit on GitHub

SciPy.jl

Build Status Stable Dev

A Julia interface for SciPy using PyCall.jl.

You can use many useful scientific functions of SciPy from Julia codes.

You can know which kind of functions are available in the SciPy Reference Guide.

Requirements

Julia 1.4.x or higher.

Install

The package is tested against, and being developed for, Julia 1.4 and above on Linux, macOS, and Windows.

You can install this package with:

using Pkg;Pkg.add("SciPy")

and then just import it with using SciPy.

If you want use latest development code, clone this repo.

Then, command this in Pkg mode:

Pkg> dev /path/to/module

How to use

You can access each SciPy modules as belows:

scipy.cluster

SciPy.cluster — Constant

scipy.cluster module

Examples

You can do k-means clustering using this module:

julia> features  = [[ 1.9  2.3];
                    [ 1.5 2.5];
                    [ 0.8 0.6];
                    [ 0.4 1.8];
                    [ 0.1 0.1];
                    [ 0.2 1.8];
                    [ 2.0 0.5];
                    [ 0.3 1.5];
                    [ 1.0 1.0]]
9×2 Array{Float64,2}:
 1.9  2.3
 1.5  2.5
 0.8  0.6
 0.4  1.8
 0.1  0.1
 0.2  1.8
 2.0  0.5
 0.3  1.5
 1.0  1.0

julia> whitened = SciPy.cluster.vq.whiten(features)
9×2 Array{Float64,2}:
 2.7396    2.91001
 2.16284   3.16306
 1.15351   0.759134
 0.576757  2.2774
 0.144189  0.126522
 0.288379  2.2774
 2.88379   0.632612
 0.432568  1.89784
 1.44189   1.26522

julia> SciPy.cluster.vq.kmeans(whitened, [whitened[1,:] whitened[3,:]] )
([1.1174670798453024 1.8345740800894272; 2.8837860125040065 0.6326117517549749], 1.073399
3090584457)
source

scipy.constants

SciPy.constants — Constant

scipy.constants module

Examples

You can access each constants:

julia> SciPy.constants.golden
1.618033988749895

julia> SciPy.constants.physical_constants["electron mass"] 
(9.10938356e-31, "kg", 1.1e-38)

julia> SciPy.constants.convert_temperature([-40, 40.0], "Celsius", "Kelvin") 

2-element Array{Float64,1}:
 233.14999999999998
 313.15
source

scipy.fft

SciPy.fft — Constant

scipy.fft module

Examples

You can use FFT (Fast Fourier Transform):


julia> SciPy.fft.fft(exp.(π/8 * collect(1:8)))
8-element Array{Complex{Float64},1}:
   68.17385416403044 - 0.0im
   1.408601300061675 + 31.248171041435185im
 -10.268363617931092 + 15.207165888808841im
 -12.695027025520982 + 6.493878653648949im
 -13.216494113330363 - 0.0im
 -12.695027025520982 - 6.493878653648949im
 -10.268363617931092 - 15.207165888808841im
   1.408601300061675 - 31.248171041435185im
source

scipy.interpolate

scipy.io

SciPy.io — Constant

scipy.io module

Examples

You can save a MATLAB-style .mat file:

julia> mdic = Dict([("a", 100), ("label", "experiment")])
Dict{String,Any} with 2 entries:
  "label" => "experiment"
  "a"     => 100

julia> SciPy.io.savemat("sample_data.mat", mdic)
source

scipy.linalg

scipy.ndimage

SciPy.ndimage — Constant

scipy.ndimage module

Examples

You can compute a multidimensional convolution:

julia> k = [[1 1 1];[1 1 0];[1 0 0]]
3×3 Array{Int64,2}:
 1  1  1
 1  1  0
 1  0  0

julia> a = [[1 2 0 0];
            [5 3 0 4];
            [0 0 0 7];
            [9 3 0 0]]
4×4 Array{Int64,2}:
 1  2  0  0
 5  3  0  4
 0  0  0  7
 9  3  0  0

julia> SciPy.ndimage.convolve(a, k, mode="constant", cval=0.0)
4×4 Array{Int64,2}:
 11  10   7   4
 10   3  11  11
 15  12  14   7
 12   3   7   0
source

scipy.odr

SciPy.odr — Constant

scipy.odr module

Examples

You can calculate orthogonal distance regression with an exponential model:

julia> x = collect(0.0:5.0);

julia> y = -10.0 .+ exp.(0.5*x);

julia> data = SciPy.odr.Data(x, y)
PyObject <scipy.odr.odrpack.Data object at 0x7fe5fda4ccc0>

julia> data = SciPy.odr.Data(x, y);

julia> odr_obj = SciPy.odr.ODR(data, SciPy.odr.exponential);

julia> output = odr_obj.run();

julia> println(output.beta)
[-10.0, 0.5]
source

scipy.optimize

SciPy.optimize — Constant

scipy.optimize module

Examples

You can optimize the Rosen function:

julia> x0 = [1.3, 0.7, 0.8, 1.9, 1.2];

julia> res = SciPy.optimize.minimize(SciPy.optimize.rosen, x0, method="Nelder-Mead", tol=1e-6)
Dict{Any,Any} with 8 entries:
  "fun"           => 1.94206e-13
  "nit"           => 295
  "nfev"          => 494
  "status"        => 0
  "success"       => true
  "message"       => "Optimization terminated successfully."
  "x"             => [1.0, 1.0, 1.0, 1.0, 1.0]
  "final_simplex" => ([1.0 1.0 … 1.0 1.0; 1.0 1.0 … 1.0 1.0; … ; 1.0 1.0 … 1.0 1.0; 1.0 1.0 … 1.0 1.0], [1…
source

scipy.signal

scipy.sparse

scipy.spatial

SciPy.spatial — Constant

scipy.spatial module

Examples

You can calculate several rotation representations:

julia> R = SciPy.spatial.transform.Rotation;

julia> r = R.from_quat([0, 0, sin(π/4.0), cos(π/4)]);

julia> r.as_matrix()
3×3 Array{Float64,2}:
 2.22045e-16  -1.0          0.0
 1.0           2.22045e-16  0.0
 0.0           0.0          1.0

julia> r.as_rotvec()
3-element Array{Float64,1}:
 0.0
 0.0
 1.5707963267948963

julia> r.as_euler("zyx", degrees=true)
3-element Array{Float64,1}:
 90.0
  0.0
  0.0
source

scipy.stats

Accessing docstring

You can access docstring of a SciPy function:

help?> SciPy.io.savemat

    Save a dictionary of names and arrays into a MATLAB-style .mat file.
    ...

Index

API Reference