The scitkit-learn request data dimension must match the size of the training data

Question

I am trying to use this code on a scikit learning site:

http://scikit-learn.org/stable/auto_examples/classification/plot_classifier_comparison.html

I use my data. My problem is that I have a lot more than two functions. If I want to "expand" the functions from 2 to 3 or 4 ....

I get:

"the size of the request data should be the size of the training data

def machine(): with open("test.txt",'r') as csvr: reader= csv.reader(csvr,delimiter='\t') for i,row in enumerate(reader): if i==0: pass elif '' in row[2:]: pass else: liste.append(map(float,row[2:])) a = np.array(liste) h = .02 names = ["Nearest Neighbors", "Linear SVM", "RBF SVM", "Decision Tree", "Random Forest", "AdaBoost", "Naive Bayes", "LDA", "QDA"] classifiers = [ KNeighborsClassifier(1), SVC(kernel="linear", C=0.025), SVC(gamma=2, C=1), DecisionTreeClassifier(max_depth=5), RandomForestClassifier(max_depth=5, n_estimators=10, max_features=1), AdaBoostClassifier(), GaussianNB(), LDA(), QDA()] X = a[:,:3] y = np.ravel(a[:,13]) linearly_separable = (X, y) datasets =[linearly_separable] figure = plt.figure(figsize=(27, 9)) i = 1 for ds in datasets: X, y = ds X = StandardScaler().fit_transform(X) X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=.4) x_min, x_max = X[:, 0].min() - .5, X[:, 0].max() + .5 y_min, y_max = X[:, 1].min() - .5, X[:, 1].max() + .5 xx, yy = np.meshgrid(np.arange(x_min, x_max, h), np.arange(y_min, y_max, h)) cm = plt.cm.RdBu cm_bright = ListedColormap(['#FF0000', '#0000FF']) ax = plt.subplot(len(datasets), len(classifiers) + 1, i) ax.scatter(X_train[:, 0], X_train[:, 1], c=y_train, cmap=cm_bright) ax.scatter(X_test[:, 0], X_test[:, 1], c=y_test, cmap=cm_bright, alpha=0.6) ax.set_xlim(xx.min(), xx.max()) ax.set_ylim(yy.min(), yy.max()) ax.set_xticks(()) ax.set_yticks(()) i += 1 for name, clf in zip(names, classifiers): ax = plt.subplot(len(datasets), len(classifiers) + 1, i) print clf.fit(X_train, y_train) score = clf.score(X_test, y_test) print y.shape, X.shape if hasattr(clf, "decision_function"): Z = clf.decision_function(np.c_[xx.ravel(), yy.ravel()]) print Z else: Z = clf.predict_proba(np.c_[xx.ravel(), yy.ravel()])[:, 1] Z = Z.reshape(xx.shape) ax.contourf(xx, yy, Z, cmap=cm, alpha=.8) ax.scatter(X_train[:, 0], X_train[:, 1], c=y_train, cmap=cm_bright) ax.scatter(X_test[:, 0], X_test[:, 1], c=y_test, cmap=cm_bright, alpha=0.6) ax.set_xlim(xx.min(), xx.max()) ax.set_ylim(yy.min(), yy.max()) ax.set_xticks(()) ax.set_yticks(()) ax.set_title(name) ax.text(xx.max() - .3, yy.min() + .3, ('%.2f' % score).lstrip('0'), size=15, horizontalalignment='right') i += 1 figure.subplots_adjust(left=.02, right=.98) plt.show() 

In this case, I use three functions. What am I doing wrong in the code, is it something with the X_train and X_test data? Only two functions, everything is in order.

my X value:

 (array([[ 1., 1., 0.], [ 1., 0., 0.], [ 1., 0., 0.], [ 1., 0., 0.], [ 1., 1., 0.], [ 1., 0., 0.], [ 1., 0., 0.], [ 3., 3., 0.], [ 1., 1., 0.], [ 1., 1., 0.], [ 0., 0., 0.], [ 0., 0., 0.], [ 0., 0., 0.], [ 0., 0., 0.], [ 0., 0., 0.], [ 0., 0., 0.], [ 4., 4., 2.], [ 0., 0., 0.], [ 6., 3., 0.], [ 5., 3., 2.], [ 2., 2., 0.], [ 4., 4., 2.], [ 2., 1., 0.], [ 2., 2., 0.]]), array([ 1., 1., 1., 1., 0., 1., 1., 0., 1., 1., 0., 1., 1., 1., 1., 1., 0., 1., 1., 0., 1., 0., 1., 1.])) 

The first array is an array of X, and the second array is an array of y (target).

I apologize for the wrong format = error:

  Traceback (most recent call last): File "allM.py", line 144, in <module> mainplot(namePlot,1,2) File "allM.py", line 117, in mainplot Z = clf.predict_proba(np.c_[xx.ravel(), yy.ravel()])[:, 1] File "/usr/local/lib/python2.7/dist-packages/sklearn/neighbors/classification.py", line 191, in predict_proba neigh_dist, neigh_ind = self.kneighbors(X) File "/usr/local/lib/python2.7/dist-packages/sklearn/neighbors/base.py", line 332, in kneighbors return_distance=return_distance) File "binary_tree.pxi", line 1298, in sklearn.neighbors.kd_tree.BinaryTree.query (sklearn/neighbors/kd_tree.c:10433) ValueError: query data dimension must match training data dimension 

and this is an X array without putting it in the ds dataset.

 [[ 1. 1. 0.][ 1. 0. 0.][ 1. 0. 0.][ 1. 0. 0.][ 1. 1. 0.][ 1. 0. 0.][ 1. 0. 0.][ 3. 3. 0.][ 1. 1. 0.][ 1. 1. 0.][ 0. 0. 0.][ 0. 0. 0.][ 0. 0. 0.][ 0. 0. 0.][ 0. 0. 0.][ 0. 0. 0.][ 4. 4. 2.][ 0. 0. 0.][ 6. 3. 0.][ 5. 3. 2.][ 2. 2. 0.][ 4. 4. 2.][ 2. 1. 0.][ 2. 2. 0.]] 

Answer 1

This is because clf.predict_proba() requires an array in which each row has the same number of elements as the rows in the training data, in other words, input with the form (num_rows, 3) .

When you worked with two-dimensional examples, it worked because the result of np.c_[xx.ravel(), yy.ravel()] is an array with two-element strings:

 print np.c_[xx.ravel(), yy.ravel()].shape (45738, 2) 

These instances have two elements because they are created using np.meshgrid , which is used in the sample code to create a set of inputs to cover two-dimensional space that will be well drawn. Try passing an array with three elements to clf.predict_proba , and everything should work fine.

If you want to reproduce this specific piece of example code, you will need to create a 3D mesh as described in this SO question. You can also display the results in 3D, where mplot3d will serve as a good starting point, although based on (admittedly short) looked at what I gave on the graph in the sample code, I suspect that this may be more of a problem than it costs. I'm not quite sure what the 3D counterpart of these plots looks like.