I am confused as what's wrong and how to approach fixing it:
def plot_boundary(my_svm, xmin, xmax, ymin, ymax):
"""
Function to plot the decision boundary for a trained SVM
It works by making a grid of x1 ("xvals") and x2 ("yvals") points,
And for each, compute whether the SVM classifies that point as
True or False. Then, a contour is drawn with a built-in pyplot function.
"""
# Hint: Use the following method built into the svm module
# my_svm.predict(np.array([x1[i],x2[j]]))
#np.array([xvals[x]), np.array(yvals[y]])
h = 0.02
#xx = np.linspace(xmin, xmax, 0.02)
#yy = np.linspace(ymin, ymax, 0.02)
xx, yy = np.meshgrid(np.arange(xmin, xmax, h), np.arange(ymin, ymax, h))
#xx, yy = np.mgrid[xmin:xmax:200j, ymin:ymax:200j]
Z = my_svm.predict(np.c_[xx.ravel(), yy.ravel()])
#Z = my_svm.predict(np.array(Xval[xx]), np.array(yvals[yy]))
#Z = my_svm.predict(Xval, yval)
#Z = my_svm.decision_function(np.c_[xx.ravel(), yy.ravel()])
# Put the result into a color plot
Z = Z.reshape(xx.shape)
#plt.contour(xx, yy, Z)
color_map = plt.cm.RdBu
#ax.contourf(xx, yy, Z, cmap=color_map, alpha=.8)
plt.contour(xx, yy, Z, colors=['k', 'k', 'k'], linestyles=['--', '-', '--'], levels=[-.5, 0, .5])
plt.show(block=False)
#my_svm.predict(np.array([Xvals[x],yvals[y]]))
#contour = plt.contour(x1,x2, mtrx , levels=[0] )
#contour = plt.contour( )
plt.title("Decision Boundary")
and then I call it with:
plot_data()
plot_boundary(clf,-.5,.3,-.8,.6)
In which svm is:
from sklearn import svm
def gauss_kernel(x1, x2, gamma):
x1 = x1.flatten()
x2 = x2.flatten()
sigma = math.sqrt(gamma)
return np.exp(-np.sum((x1-x2)**2)/(2*sigma**2))
# from @lejlot http://stackoverflow.com/a/26962861/583834
def gaussianKernelGramMatrix(X1, X2, K_function=gauss_kernel, gamma=0.1):
"""(Pre)calculates Gram Matrix K"""
gram_matrix = np.zeros((X1.shape[0], X2.shape[0]))
for i, x1 in enumerate(X1):
for j, x2 in enumerate(X2):
gram_matrix[i, j] = K_function(x1, x2, gamma)
return gram_matrix
gamma=0.1
y = y.flatten()
clf = svm.SVC(kernel="precomputed", verbose=2, C=2.0)
clf.fit(gaussianKernelGramMatrix(X,X, gauss_kernel, gamma=gamma), y)
As you see I have tried various combination of achieving xx as well as trying both predict method as well as decision_function methods in plot_boundary. I wonder why both time I get the same error and how could that be fixed? [sklearn noob here]
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
<ipython-input-145-61aad96e14ed> in <module>()
1 plot_data()
----> 2 plot_boundary(clf,-.5,.3,-.8,.6)
<ipython-input-144-1fdc3948d9d2> in plot_boundary(my_svm, xmin, xmax, ymin, ymax)
15 #xx, yy = np.mgrid[xmin:xmax:200j, ymin:ymax:200j]
16
---> 17 Z = my_svm.predict(np.c_[xx.ravel(), yy.ravel()])
18 #Z = my_svm.predict(np.array(Xval[xx]), np.array(yvals[yy]))
19 #Z = my_svm.predict(Xval, yval)
~/anaconda/lib/python3.6/site-packages/sklearn/svm/base.py in predict(self, X)
546 Class labels for samples in X.
547 """
--> 548 y = super(BaseSVC, self).predict(X)
549 return self.classes_.take(np.asarray(y, dtype=np.intp))
550
~/anaconda/lib/python3.6/site-packages/sklearn/svm/base.py in predict(self, X)
306 y_pred : array, shape (n_samples,)
307 """
--> 308 X = self._validate_for_predict(X)
309 predict = self._sparse_predict if self._sparse else self._dense_predict
310 return predict(X)
~/anaconda/lib/python3.6/site-packages/sklearn/svm/base.py in _validate_for_predict(self, X)
453 raise ValueError("X.shape[1] = %d should be equal to %d, "
454 "the number of samples at training time" %
--> 455 (X.shape[1], self.shape_fit_[0]))
456 elif n_features != self.shape_fit_[1]:
457 raise ValueError("X.shape[1] = %d should be equal to %d, "
ValueError: X.shape[1] = 2 should be equal to 211, the number of samples at training time
