void DataProcess::ReadFileWidBucket(const char * fname, int fcol,
double dw, vector<int>& _dbRes) {
vector<double> data;
ReadFileDouble(fname, fcol, data);
Bucketized(data, dw, _dbRes);
}
void DataProcess::ReadFileBucket(const char * fname, int fcol,
int bukNum, vector<int> _dbRes) {
vector<double> data;
ReadFileDouble(fname, fcol,data);
double dw = getBukWidth(bukNum);
Bucketized(data, dw,_dbRes);
}
void DataProcess::z_normalizeData(string inputFilename, int fcol, string outputFilename, double rangeMultipler){
vector<double> data;
data.clear();
cout<<"start processing file:"<<inputFilename<<" col:"<<fcol<<endl;
ReadFileDouble(inputFilename.c_str(), fcol, data);
double sum=0,sumSqau=0;
for(size_t i=0;i<data.size();i++){
sum+=data[i];
sumSqau+=(data[i]*data[i]);
}
double mean = sum/data.size();
double dev = sqrt(sumSqau/data.size() - mean*mean);
string line;
ifstream ifile(inputFilename.c_str());
ofstream outf;
outf.open(outputFilename.c_str(), ios::out);
outf<<"#mean:"<<mean<<" standard variance:"<<dev<<endl;
if (ifile.is_open()) {
while (getline(ifile, line)) {
std::size_t found = line.find("#");
if(found!=std::string::npos) {
outf<<line<<endl;
continue;
}
vector<string> nstring = split(line, ",");
string myString = eraseSpace(nstring[fcol]);
double value = (atof)(myString.c_str());
//cout<<"my string"<<myString<<endl;
for(int i=0;i<nstring.size();i++){
if(i!=fcol){
outf<<nstring[i]<<",";
}else{
outf<< (value-mean)/dev*rangeMultipler<<",";
}
}
outf<<endl;
}
}
cout<<"output normalized data in:"<<outputFilename<<endl;
outf.close();
}
/**
* _data:output result
*/
void DataProcess::ReadFileFloat(const char* fname, vector<vector<float> >& _data, int lineNum){
vector<vector<double> > data_db;
ReadFileDouble(fname, data_db);
_data.resize(data_db.size());
for(int i=0;i<data_db.size();i++){
_data[i].resize(data_db[i].size());
for(int j=0;j<data_db[i].size();j++){
_data[i][j] = (float) data_db[i][j];
}
}
}
void DataProcess::ReadFileFloat(const char* fname,int fcol, vector<float>& _data){
vector<double> df;
ReadFileDouble(fname,fcol,df);
_data.resize(df.size());
for(int i=0;i<df.size();i++){
_data[i] = (float)(df[i]);
}
}
int main(int argc, char** argv)
{
double *fv1=NULL;
double *fv2=NULL;
double distance=0;
if (argc != 3) {
fprintf(stderr,"usage: spytec_distance <fv1_path> <fv2_path>\n");
exit(-1);
}
fv1 = ReadFileDouble(argv[1], QTD_COEFS);
fv2 = ReadFileDouble(argv[2], QTD_COEFS);
distance=L2DoubleDistance(fv1, fv2, QTD_COEFS);
printf("%lf\n",distance);
free(fv1);
free(fv2);
return(0);
}
void DataProcess::z_normalizeData(string inputFilename, int fcol_start,int fcol_end, string outputFilename,double rangeMultipler){
vector<double> mean(fcol_end-fcol_start+1,0);
vector<double> dev(fcol_end-fcol_start+1,0);
for(int i=fcol_start;i<=fcol_end;i++){
vector<double> data;
ReadFileDouble(inputFilename.c_str(), i,data);
z_normalizeData_perVector(data, mean[i-fcol_start],dev[i-fcol_start]);
}
string line;
ifstream ifile(inputFilename.c_str());
ofstream outf;
outf.open(outputFilename.c_str(), ios::out);
if (ifile.is_open()) {
while (getline(ifile, line)) {
std::size_t found = line.find("#");
if (found != std::string::npos) {
outf << line << endl;
continue;
}
vector<string> nstring = split(line, ",");
for (int i = 0; i < nstring.size(); i++) {
if (i < fcol_start||i>fcol_end) {
outf << nstring[i] << ",";
} else {
string myString = eraseSpace(nstring[i]);
double value = (atof)(myString.c_str());
outf << (value - mean[i-fcol_start]) / dev[i-fcol_start] * rangeMultipler << ",";
}
}
outf << endl;
}
}
}