Merge pull request #43 from haoxingz/magic_test

Magic test

Author: arokem

pymatbridge/matlab_magic.py

@@ -28,7 +28,7 @@
     has_io = False
     no_io_str = "Must have h5py and scipy.io to perform i/o"
     no_io_str += "operations with the Matlab session"
-    
+
 from IPython.core.displaypub import publish_display_data
 from IPython.core.magic import (Magics, magics_class, cell_magic, line_magic,
                                 line_cell_magic, needs_local_scope)
@@ -39,7 +39,7 @@
 
 import pymatbridge as pymat
 
-    
+
 class MatlabInterperterError(RuntimeError):
     """
     Some error occurs while matlab is running
@@ -52,7 +52,7 @@ def __unicode__(self):
         s = "Failed to parse and evaluate line %r.\n Matlab error message: %r"%\
                 (self.line, self.err)
         return s
-    
+
     if PY3:
         __str__ = __unicode__
     else:
@@ -66,25 +66,54 @@ def loadmat(fname):
     """
 
     f = h5py.File(fname)
-    data = f.values()[0][:]
-    if len(data.dtype) > 0:
-        # must be complex data
-        data = data['real'] + 1j * data['imag']        
-    return data
+
+    for var_name in f.iterkeys():
+        if isinstance(f[var_name], h5py.Dataset):
+            # Currently only supports numerical array
+            data = f[var_name].value
+            if len(data.dtype) > 0:
+                # must be complex data
+                data = data['real'] + 1j * data['imag']
+            return np.squeeze(data.T)
+
+        elif isinstance(f[var_name], h5py.Group):
+            data = {}
+            for mem_name in f[var_name].iterkeys():
+                if isinstance(f[var_name][mem_name], h5py.Dataset):
+                    # Check if the dataset is a string
+                    attr = h5py.AttributeManager(f[var_name][mem_name])
+                    if (attr.__getitem__('MATLAB_class') == 'char'):
+                        is_string = True
+                    else:
+                        is_string = False
+
+                    data[mem_name] = f[var_name][mem_name].value
+                    data[mem_name] = np.squeeze(data[mem_name].T)
+
+                    if is_string:
+                        result = ''
+                        for asc in data[mem_name]:
+                            result += chr(asc)
+                        data[mem_name] = result
+                else:
+                    # Currently doesn't support nested struct
+                    pass
+
+            return data
 
 
 def matlab_converter(matlab, key):
     """
 
     Reach into the matlab namespace and get me the value of the variable
-    
+
     """
     tempdir = tempfile.gettempdir()
     # We save as hdf5 in the matlab session, so that we can grab large
     # variables:
     matlab.run_code("save('%s/%s.mat','%s','-v7.3')"%(tempdir, key, key),
                     maxtime=matlab.maxtime)
-    
+
     return loadmat('%s/%s.mat'%(tempdir, key))
 
 
@@ -113,17 +142,17 @@ def __init__(self, shell,
         maxtime : float
            The maximal time to wait for responses for matlab (in seconds).
            Default: 10 seconds.
-           
+
         pyconverter : callable
             To be called on matlab variables returning into the ipython
             namespace
-            
+
         matlab_converter : callable
-            To be called on values in ipython namespace before 
+            To be called on values in ipython namespace before
             assigning to variables in matlab.
 
         cache_display_data : bool
-            If True, the published results of the final call to R are 
+            If True, the published results of the final call to R are
             cached in the variable 'display_cache'.
 
         """
@@ -133,7 +162,7 @@ def __init__(self, shell,
         self.Matlab = pymat.Matlab(matlab, maxtime=maxtime)
         self.Matlab.start()
         self.pyconverter = pyconverter
-        self.matlab_converter = matlab_converter        
+        self.matlab_converter = matlab_converter
 
     def __del__(self):
         """shut down the Matlab server when the object dies.
@@ -154,9 +183,9 @@ def eval(self, line):
         if run_dict['success'] == 'false':
             raise MatlabInterperterError(line, run_dict['content']['stdout'])
 
-        # This is the matlab stdout: 
+        # This is the matlab stdout:
         return run_dict
-        
+
     @magic_arguments()
     @argument(
         '-i', '--input', action='append',
@@ -180,7 +209,7 @@ def matlab(self, line, cell=None, local_ns=None):
         """
 
         Execute code in matlab
-        
+
         """
         args = parse_argstring(self.matlab, line)
 
@@ -210,7 +239,7 @@ def matlab(self, line, cell=None, local_ns=None):
                     except KeyError:
                         val = self.shell.user_ns[input]
                     # We save these input arguments into a .mat file:
-                    tempdir = tempfile.gettempdir() 
+                    tempdir = tempfile.gettempdir()
                     sio.savemat('%s/%s.mat'%(tempdir, input),
                                 eval("dict(%s=val)"%input), oned_as='row')
 
@@ -219,7 +248,7 @@ def matlab(self, line, cell=None, local_ns=None):
 
             else:
                 raise RuntimeError(no_io_str)
-            
+
         text_output = ''
         #imgfiles = []
 
@@ -234,14 +263,14 @@ def matlab(self, line, cell=None, local_ns=None):
             e_s += "\n-----------------------"
             e_s += "\nAre you sure Matlab is started?"
             raise RuntimeError(e_s)
-            
-        
+
+
 
         text_output += result_dict['content']['stdout']
         # Figures get saved by matlab in reverse order...
         imgfiles = result_dict['content']['figures'][::-1]
         data_dir = result_dict['content']['datadir']
-        
+
         display_data = []
         if text_output:
             display_data.append(('MatlabMagic.matlab',
@@ -251,7 +280,7 @@ def matlab(self, line, cell=None, local_ns=None):
             if len(imgf):
                 # Store the path to the directory so that you can delete it
                 # later on:
-                image = open(imgf, 'rb').read() 
+                image = open(imgf, 'rb').read()
                 display_data.append(('MatlabMagic.matlab',
                                      {'image/png': image}))
 
@@ -261,24 +290,24 @@ def matlab(self, line, cell=None, local_ns=None):
         # Delete the temporary data files created by matlab:
         if len(data_dir):
             rmtree(data_dir)
-        
+
         if args.output:
             if has_io:
                 for output in ','.join(args.output).split(','):
                     self.shell.push({output:self.matlab_converter(self.Matlab,
                                                               output)})
             else:
                 raise RuntimeError(no_io_str)
-                
-            
+
+
 _loaded = False
 def load_ipython_extension(ip, **kwargs):
     """Load the extension in IPython."""
     global _loaded
     if not _loaded:
         ip.register_magics(MatlabMagics(ip, **kwargs))
         _loaded = True
-        
+
 def unload_ipython_extension(ip):
     global _loaded
     if _loaded:

pymatbridge/tests/test_magic.py

@@ -0,0 +1,83 @@
+import pymatbridge as pymat
+import IPython
+
+import numpy.testing as npt
+
+class TestMagic:
+
+    # Create an IPython shell and load Matlab magic
+    @classmethod
+    def setup_class(cls):
+        cls.ip = IPython.InteractiveShell()
+        cls.ip.run_cell('import random')
+        cls.ip.run_cell('import numpy as np')
+        pymat.load_ipython_extension(cls.ip)
+
+    # Unload the magic, shut down Matlab
+    @classmethod
+    def teardown_class(cls):
+        pymat.unload_ipython_extension(cls.ip)
+
+
+    # Test single operation on different data structures
+    def test_cell_magic_number(self):
+        # A double precision real number
+        self.ip.run_cell("a = np.float64(random.random())")
+        self.ip.run_cell_magic('matlab', '-i a -o b', 'b = a*2;')
+        npt.assert_almost_equal(self.ip.user_ns['b'],
+                                self.ip.user_ns['a']*2, decimal=7)
+
+        # A complex number
+        self.ip.run_cell("x = 3.34+4.56j")
+        self.ip.run_cell_magic('matlab', '-i x -o y', 'y = x*(11.35 - 23.098j)')
+        self.ip.run_cell("res = x*(11.35 - 23.098j)")
+        npt.assert_almost_equal(self.ip.user_ns['y'],
+                                self.ip.user_ns['res'], decimal=7)
+
+
+    def test_cell_magic_array(self):
+        # Random array multiplication
+        self.ip.run_cell("val1 = np.random.random_sample((3,3))")
+        self.ip.run_cell("val2 = np.random.random_sample((3,3))")
+        self.ip.run_cell("respy = np.dot(val1, val2)")
+        self.ip.run_cell_magic('matlab', '-i val1,val2 -o resmat',
+                               'resmat = val1 * val2')
+        npt.assert_almost_equal(self.ip.user_ns['resmat'],
+                                self.ip.user_ns['respy'], decimal=7)
+
+
+    def test_line_magic(self):
+        # Some operation in Matlab
+        self.ip.run_line_magic('matlab', 'a = [1 2 3]')
+        self.ip.run_line_magic('matlab', 'res = a*2')
+        # Get the result back to Python
+        self.ip.run_cell_magic('matlab', '-o actual', 'actual = res')
+
+        self.ip.run_cell("expected = np.array([2, 4, 6])")
+        npt.assert_almost_equal(self.ip.user_ns['actual'],
+                                self.ip.user_ns['expected'], decimal=7)
+
+    def test_figure(self):
+        # Just make a plot to get more testing coverage
+        self.ip.run_line_magic('matlab', 'plot([1 2 3])')
+
+
+    def test_matrix(self):
+        self.ip.run_cell("in_array = np.array([[1,2,3], [4,5,6]])")
+        self.ip.run_cell_magic('matlab', '-i in_array -o out_array',
+                               'out_array = in_array;')
+        npt.assert_almost_equal(self.ip.user_ns['out_array'],
+                                self.ip.user_ns['in_array'],
+                                decimal=7)
+
+    # Matlab struct type should be converted to a Python dict
+    def test_struct(self):
+        self.ip.run_cell('num = 2.567')
+        self.ip.run_cell('num_array = np.array([1.2,3.4,5.6])')
+        self.ip.run_cell('str = "Hello World"')
+        self.ip.run_cell_magic('matlab', '-i num,num_array,str -o obj',
+                                'obj.num = num;obj.num_array = num_array;obj.str = str;')
+        npt.assert_equal(isinstance(self.ip.user_ns['obj'], dict), True)
+        npt.assert_equal(self.ip.user_ns['obj']['num'], self.ip.user_ns['num'])
+        npt.assert_equal(self.ip.user_ns['obj']['num_array'], self.ip.user_ns['num_array'])
+        npt.assert_equal(self.ip.user_ns['obj']['str'], self.ip.user_ns['str'])