af286ba5280614ecc8371b71ed53cc6447d1183b,GPy/kern/src/static.py,Precomputed,K,#Precomputed#Any#Any#,231

Before Change

        
        super(Precomputed, self).__init__(input_dim, covariance_matrix, variance, active_dims, name)
    def K(self, X, X2=None):
        if X2 is None:
            return self.variance * self.fixed_K[X[:,0].astype("int")][:,X[:,0].astype("int")]
        else:
            return self.variance * self.fixed_K[X[:,0].astype("int")][:,X2[:,0].astype("int")]

    def Kdiag(self, X):
        return self.variance * self.fixed_K[X[:,0].astype("int")][:,X[:,0].astype("int")].diagonal()

    def update_gradients_full(self, dL_dK, X, X2=None):

After Change

        return self.fixed_K[i1,:][:,i2]

    def K(self, X, X2=None):
        return self.variance * self._index(X, X2)

    def Kdiag(self, X):
        return self.variance * self._index(X,None).diagonal()

Instances