From 063d7767d4c3623fd7376ba8a844df630e0dae80 Mon Sep 17 00:00:00 2001
From: maedoc <marmaduke.woodman@univ-amu.fr>
Date: Fri, 18 Sep 2015 17:05:47 +0200
Subject: [PATCH] add numpy to matlab array converter

---
 conda.m |  2 +-
 np2m.m  | 62 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 63 insertions(+), 1 deletion(-)
 create mode 100644 np2m.m

diff --git a/conda.m b/conda.m
index f7629d1..a67b12c 100644
--- a/conda.m
+++ b/conda.m
@@ -2,8 +2,8 @@ function conda(cmd, varargin)
 
     persistent conda_path conda_env conda_py
     
-    
     switch cmd
+        % TODO this prevents true conda install from working.
         case 'install'
             install(varargin{:});
         case 'use'
diff --git a/np2m.m b/np2m.m
new file mode 100644
index 0000000..1f13ff8
--- /dev/null
+++ b/np2m.m
@@ -0,0 +1,62 @@
+function y = np2m(x)
+% Convert NumPy array to MATLAB array
+
+% kudos to http://fr.mathworks.com/matlabcentral/answers/157347-convert-python-numpy-array-to-double#answer_157736
+
+%% setup table of conversions from Python array type codes to MATLAB classes
+% is persistent faster than setting up the table every time?
+persistent t2m % type code to matlab class
+
+if isempty(t2m)
+    t2m.c = @char;
+    t2m.b = @int8;
+    t2m.B = @uint8;
+    t2m.h = @int16;
+    t2m.H = @int16;
+    t2m.i = @int32;
+    t2m.I = @uint32;
+    t2m.l = @int64;
+    t2m.L = @uint64;
+    t2m.f = @single;
+    t2m.d = @double;
+end
+
+%% handle type of array
+type_code = char(x.dtype.char);
+matlab_class = t2m.(type_code);
+
+%% convert data
+y = matlab_class(py.array.array(type_code, py.numpy.nditer(x)));
+
+%% handle shape if required
+ndim = x.ndim * 1;
+if ndim > 1
+    shape = int32(zeros([1 ndim]));
+    for i=1:ndim
+        shape(i) = x.shape{ndim+1-i}*1;
+    end
+    y = reshape(y, shape);
+end
+
+%% benchmarks
+% Running for various sizes, we can see it's not zero-copy, so to be used
+% with this in mind.
+
+%{
+[np2m] 0.08 MB, 0.04 ms / iter / KB
+[np2m] 0.76 MB, 0.03 ms / iter / KB
+[np2m] 7.63 MB, 0.03 ms / iter / KB
+%}
+
+if 0
+    %%
+    n = 1000000;
+    m = 100;
+    x = py.numpy.random.randn(n).reshape([10, -1]);
+    tic
+    for i=1:m
+        y = np2m(x);
+    end
+    fprintf('[np2m] %0.2f MB, %.2f ms / iter / KB\n', ...
+        n*8/1024/1024, toc*1000/m/(n*8/1024));
+end
\ No newline at end of file