src/test/scripts/applications/nn/component/logcosh.dml

#-------------------------------------------------------------
#
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements.  See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership.  The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License.  You may obtain a copy of the License at
#
#   http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied.  See the License for the
# specific language governing permissions and limitations
# under the License.
#
#-------------------------------------------------------------

source("scripts/nn/layers/logcosh_loss.dml") as logcosh

logcosh_test_1 = function(){
    X = rand(rows = 3, cols = 3, pdf = "normal")
    labels = X 
    loss = logcosh::forward(X, labels)
    if( loss > 1e-10)
        print("ERROR: Loss of perfect predictions should be equal to 0 but was " + loss)
}

logcosh_test_2 = function(){
    X = rand(rows = 3, cols = 1, pdf = "normal")
    labels_1 = X + 1
    labels_2 = X - 1 
    loss_1 = logcosh::forward(X, labels_1)
    loss_2 = logcosh::forward(X, labels_2)
    if( loss_1 - loss_2 > 1e-10)
        print("ERROR: Loss for values predicted above and below by same margin should be exactly the same loss " + loss_1 +  " : " + loss_2)
}

logcosh_test_3 = function(){
    X = rand(rows = 6, cols = 2, pdf = "normal")
    labels_1 = X + 1
    labels_2 = X - 1 
    gradient_1 = sum(logcosh::backward(X, labels_1))
    gradient_2 = sum(logcosh::backward(X, labels_2))
    if(gradient_1 > 0)
        print("ERROR: gradients should point negatively if prediction is to high")
    if(gradient_2 < 0)
        print("ERROR: gradients should point positivly if prediction is to low")
}

logcosh_test_1()
logcosh_test_2()
logcosh_test_3()