matrix1d sum(matrix2d myVector) {
matrix1d mySum(myVector.size(),0.0);
for (int i=0; i<myVector.size(); i++) {
for (int j=0; j<myVector[i].size(); j++) {
mySum[i]+=myVector[i][j];
}
}
return mySum;
}
matrix1d mean2(matrix2d myVector) {
matrix1d myMean(myVector[0].size(),0.0);
for (int i=0; i<myVector.size(); i++) {
for (int j=0; j<myVector[i].size(); j++) {
myMean[j]+=myVector[i][j]/double(myVector.size());
}
}
return myMean;
}
matrix1d mean1(matrix2d myVector) {
matrix1d myMean(myVector.size(),0.0);
for (int i=0; i<myVector.size(); i++) {
for (int j=0; j<myVector[i].size(); j++) {
myMean[i]+=myVector[i][j];
}
myMean[i]/=double(myVector[i].size());
}
return myMean;
}
matrix1d NeuroEvo::getAction(matrix2d state) {
matrix1d stateSum(state[0].size(), 0.0);
// state[type][state_element] -- specifies combination for state
for (size_t i = 0; i < state.size(); i++) {
for (size_t j = 0; j < state[i].size(); j++) {
stateSum[j] += state[i][j];
}
}
return getAction(stateSum);
}