ó ‡ˆ\c@sødZddlZddlmZd„Zd„Zd„Zd„Z d„Z ied 6ed 6ed 6e d 6e d 6Z d„Z d„Z d„Zd„Zie d 6ed 6e d 6ed 6Zd„Zd„Zd„Zied6ed6ed6ZdS(s)Utilities for the neural network modules iÿÿÿÿN(texpitcCs|S(sjSimply return the input array. Parameters ---------- X : {array-like, sparse matrix}, shape (n_samples, n_features) Data, where n_samples is the number of samples and n_features is the number of features. Returns ------- X : {array-like, sparse matrix}, shape (n_samples, n_features) Same as the input data. ((tX((s;lib/python2.7/site-packages/sklearn/neural_network/_base.pytidentity scCst|d|ƒS(s#Compute the logistic function inplace. Parameters ---------- X : {array-like, sparse matrix}, shape (n_samples, n_features) The input data. Returns ------- X_new : {array-like, sparse matrix}, shape (n_samples, n_features) The transformed data. tout(tlogistic_sigmoid(R((s;lib/python2.7/site-packages/sklearn/neural_network/_base.pytlogistics cCstj|d|ƒS(s)Compute the hyperbolic tan function inplace. Parameters ---------- X : {array-like, sparse matrix}, shape (n_samples, n_features) The input data. Returns ------- X_new : {array-like, sparse matrix}, shape (n_samples, n_features) The transformed data. R(tnpttanh(R((s;lib/python2.7/site-packages/sklearn/neural_network/_base.pyR-s cCs,tj|dtj|jƒjd|ƒ|S(s0Compute the rectified linear unit function inplace. Parameters ---------- X : {array-like, sparse matrix}, shape (n_samples, n_features) The input data. Returns ------- X_new : {array-like, sparse matrix}, shape (n_samples, n_features) The transformed data. iR(Rtcliptfinfotdtypetmax(R((s;lib/python2.7/site-packages/sklearn/neural_network/_base.pytrelu=s (cCsi||jddƒdd…tjf}tj|d|ƒ||jddƒdd…tjf:}|S(s(Compute the K-way softmax function inplace. Parameters ---------- X : {array-like, sparse matrix}, shape (n_samples, n_features) The input data. Returns ------- X_new : {array-like, sparse matrix}, shape (n_samples, n_features) The transformed data. taxisiNR(R Rtnewaxistexptsum(Rttmp((s;lib/python2.7/site-packages/sklearn/neural_network/_base.pytsoftmaxNs ))RRRR RcCsdS(s„Apply the derivative of the identity function: do nothing. Parameters ---------- Z : {array-like, sparse matrix}, shape (n_samples, n_features) The data which was output from the identity activation function during the forward pass. delta : {array-like}, shape (n_samples, n_features) The backpropagated error signal to be modified inplace. N((tZtdelta((s;lib/python2.7/site-packages/sklearn/neural_network/_base.pytinplace_identity_derivativefscCs||9}|d|9}dS(sóApply the derivative of the logistic sigmoid function. It exploits the fact that the derivative is a simple function of the output value from logistic function. Parameters ---------- Z : {array-like, sparse matrix}, shape (n_samples, n_features) The data which was output from the logistic activation function during the forward pass. delta : {array-like}, shape (n_samples, n_features) The backpropagated error signal to be modified inplace. iN((RR((s;lib/python2.7/site-packages/sklearn/neural_network/_base.pytinplace_logistic_derivativeus cCs|d|d9}dS(sýApply the derivative of the hyperbolic tanh function. It exploits the fact that the derivative is a simple function of the output value from hyperbolic tangent. Parameters ---------- Z : {array-like, sparse matrix}, shape (n_samples, n_features) The data which was output from the hyperbolic tangent activation function during the forward pass. delta : {array-like}, shape (n_samples, n_features) The backpropagated error signal to be modified inplace. iiN((RR((s;lib/python2.7/site-packages/sklearn/neural_network/_base.pytinplace_tanh_derivativeˆscCsd||dks,