Merge pull request #16 from yingtluo/master

Finding and plotting spread of data for one subject's run

code/utils/plot.py

@@ -0,0 +1,88 @@
+import numpy as np
+import matplotlib.pyplot as plt
+
+
+def plot_slice(vol, m, color='gray', ipn='nearest'):
+    """
+	Plots the (m+1)th slice over the 3rd dimension across all voxels.
+
+	Parameters
+	----------
+	vol : 3-D array
+	Specifies which 3-D volume to plot.
+
+	m : number
+	Specifies which slice to plot
+
+	color : String
+	Optional input that specifies the color of the plot
+
+	ipon : String
+	Optional input that specifies how the interpolation will be conducted
+
+	Returns
+	-------
+	Nothing
+	"""
+    plt.imshow(vol[:, :, m], cmap=color, interpolation=ipn)
+
+
+def plot_central_slice(vol, color='gray', ipn='nearest'):
+    """
+	Plots the central slice over the 3rd dimension across all voxels.
+	If the 3-D volume given does not have a center, then plot_central_slice
+	plots the smaller index of the two centermost slices.
+
+	Parameters
+	----------
+	vol : 3-D array
+	Specifies which 3-D volume to plot.
+
+	color : String
+	Optional input that specifies the color of the plot 
+
+	ipn : String
+	Optional input that specifies how the interpolation will be conducted
+
+	Returns
+	-------
+	Nothing
+	"""
+    c = vol.shape[2] // 2
+    plt.imshow(vol[:, :, c], cmap=color, interpolation=ipn)
+
+
+def plot_sd(data):
+    """
+	Finds standard deviation for each volume and plots those values.
+
+	Parameters
+	----------
+	data : 4-D array
+
+	Returns 
+	-------
+	stds : 1-D array
+	List of standard deviations that were plotted
+	"""
+    stds = [np.std(data[..., i]) for i in range(data.shape[-1])]
+    plt.plot(stds)
+    return stds
+
+
+def plot_var(data):
+    """
+	Finds standard deviation for each volume and plots those values.
+
+	Parameters
+	----------
+	data : 4-D array
+
+	Returns 
+	-------
+	variances : 1-D array
+	List of variances that were plotted
+	"""
+    var = [np.var(data[..., i]) for i in range(data.shape[-1])]
+    plt.plot(var)
+    return var

code/utils/tests/test_plot.py

@@ -0,0 +1,51 @@
+from __future__ import absolute_import
+from .. import plot
+
+import numpy as np
+import matplotlib
+
+from numpy.testing import assert_almost_equal, assert_array_equal
+try:
+    from mock import patch
+except:
+    from unittest.mock import patch
+import unittest
+
+class PlotSlicesTestCase(unittest.TestCase):
+    @patch.object(matplotlib.pyplot, 'imshow')
+    def test_plot_slice(self, mock_imshow):
+        plot.plot_slice(np.array([[[1, 2], [1, 2]], [[1, 2], [1, 2]]]), 0)
+        assert_array_equal(mock_imshow.call_args[0][0],
+                           np.array([[1, 1], [1, 1]]))
+        assert_array_equal(mock_imshow.call_args[1],
+                           {'cmap': 'gray',
+                            'interpolation': 'nearest'})
+
+    @patch.object(matplotlib.pyplot, 'imshow')
+    def test_plot_central_slice(self, mock_imshow):
+        plot.plot_central_slice(np.array([[[1, 2], [1, 2]], [[1, 2], [1, 2]
+                                                                 ]]))
+        assert_array_equal(mock_imshow.call_args[0][0],
+                           np.array([[2, 2], [2, 2]]))
+        assert_array_equal(mock_imshow.call_args[1],
+                           {'cmap': 'gray',
+                            'interpolation': 'nearest'})
+
+        plot.plot_central_slice(np.array([[[1, 2], [1, 2]], [[1, 2], [1, 2]
+                                                                 ]]))
+        assert_array_equal(mock_imshow.call_args[0][0],
+                           np.array([[2, 2], [2, 2]]))
+        assert_array_equal(mock_imshow.call_args[1],
+                           {'cmap': 'gray',
+                            'interpolation': 'nearest'})
+
+class PlotSpreadTestCase(unittest.TestCase):
+    @patch.object(matplotlib.pyplot, 'plot')
+    def test_plot_sd(self, mock_pyplot):
+        plot.plot_sd(np.array([[[1, 2], [1, 2]], [[1, 2], [1, 2]]]))
+        assert_array_equal(mock_pyplot.call_args[0][0], [0.0, 0.0])
+
+    @patch.object(matplotlib.pyplot, 'plot')
+    def test_plot_var(self, mock_pyplot):
+        plot.plot_var(np.array([[[1, 2], [3, 4]], [[5, 6], [7, 8]]]))
+        assert_array_equal(mock_pyplot.call_args[0][0], [5.0, 5.0])

code/utils/tests/test_variance.py

@@ -0,0 +1,49 @@
+from __future__ import absolute_import
+from .. import variance
+
+import numpy as np
+import nibabel as nib
+
+from numpy.testing import assert_almost_equal, assert_array_equal
+import unittest
+import os
+
+
+class LoadDataTestCase(unittest.TestCase):
+    def test_load_image(self):
+        # Creating mock .nii file for testing
+        data = np.array([[[[7, 9], [7, 8]], [[1, 2], [1, 8]]],
+                         [[[2, 3], [2, 1]], [[5, 4], [4, 3]]]])
+        img = nib.Nifti1Image(data, affine=np.diag([1, 1, 1, 1]))
+        nib.save(img, 'test_data.nii')
+        # Calling load_image function and testing the loaded output
+        data = variance.load_image('test_data.nii')
+        assert_array_equal(data.shape, (2, 2, 2, 2))
+        assert_array_equal(data,
+                           np.array([[[[7, 9], [7, 8]], [[1, 2], [1, 8]]],
+                                     [[[2, 3], [2, 1]], [[5, 4], [4, 3]]]]))
+        # Removing .nii file when test ends
+        os.remove('test_data.nii')
+
+    def test_isolate_vol(self):
+        data = np.array([[[[7, 9], [7, 8]], [[1, 2], [1, 8]]],
+                         [[[2, 3], [2, 1]], [[5, 4], [4, 3]]]])
+        vol0 = variance.isolate_vol(data, 0)
+        assert_array_equal(vol0,
+                           np.array([[[7, 7], [1, 1]], [[2, 2], [5, 4]]]))
+
+        vol1 = variance.isolate_vol(data, 1)
+        assert_array_equal(vol1,
+                           np.array([[[9, 8], [2, 8]], [[3, 1], [4, 3]]]))
+
+    def test_find_sd(self):
+        data = np.array([[[[7, 9], [7, 8]], [[1, 2], [1, 8]]],
+                         [[[2, 3], [2, 1]], [[5, 4], [4, 3]]]])
+        sd = variance.find_sd(data)
+        assert_almost_equal(sd, 2.69765523)
+
+    def test_find_var(self):
+        data = np.array([[[[7, 9], [7, 8]], [[1, 2], [1, 8]]],
+                         [[[2, 3], [2, 1]], [[5, 4], [4, 3]]]])
+        var = variance.find_var(data)
+        assert_almost_equal(var, 7.27734375)

code/utils/variance.py

@@ -0,0 +1,70 @@
+import numpy as np
+import nibabel as nib
+
+
+def load_image(fname):
+    """
+	Loads the .nii file 
+
+	Parameters
+	----------
+	fname : String
+	The path to the .nii file that we wish to load and get the data of
+
+	Returns
+	-------
+	data : a 4-D array
+	"""
+    img = nib.load(fname)
+    data = img.get_data()
+    return data
+
+
+def isolate_vol(data, n):
+    """
+	Selects the (n+1)th volume from the 4-D image data array by slicing over
+	the last dimension.
+
+	Parameters
+	----------
+	data : 4-D image data array
+
+	n : number
+	Indicates which index of the last dimension to slice
+
+	Returns 
+	-------
+	voln : 3-D array
+	The (n+1)th volume
+	"""
+    return data[..., n]
+
+
+def find_sd(vol):
+    """ 
+	Finds standard deviation across all voxels for one volume
+
+	Parameters
+	----------
+	vol: The voxel which we wish to calculate the standard deviation on. 
+
+	Returns
+	-------
+	sd : number
+	"""
+    return np.std(vol)
+
+
+def find_var(vol):
+    """ 
+	Finds variance across all voxels for one volume
+
+	Parameters
+	----------
+	vol: The voxel which we wish to calculate the standard deviation on. 
+
+	Returns
+	-------
+	var : number
+	"""
+    return np.var(vol)