int main(int argc, char *argv[])
{
SimpleMatrix m;
assert (m.nrow() == 0);
assert (m.ncol() == 0);
m.resize(3, 2);
assert (m.nrow() == 3);
assert (m.ncol() == 2);
m.zero();
assert (m[0][0] == 0.0);
assert (m[2][1] == 0.0);
std::vector<double> col(3, 1.0);
m.appendCol(col);
assert (m[0][0] == 0.0);
assert (m[2][2] == 1.0);
m.resize(2,3);
std::vector<double> row(3, 2.0);
m.appendRow(row);
assert (m[0][0] == 0.0);
assert (m[2][2] == 2.0);
return 0;
}
void printMatrix(SimpleMatrix& m, PlinkInputFile& f1) {
for (int i = 0; i < m.nrow(); i++ ) {
for (int j = 0; j < m.ncol(); j++) {
int geno = (int) m[i][j];
char ref = f1.ref[j];
char alt = f1.alt[j];
switch(geno) {
case 0:
printf("%c %c\t", ref, ref);
break;
case 1:
printf("%c %c\t", ref, alt);
break;
case 2:
printf("%c %c\t", alt, alt);
break;
default:
if ( geno < 0)
printf(". .\t");
break;
printf("ERROR reading!\n");
}
}
printf("\n");
}
}
/**
* Load covariate from @param fn, using specified @param covNameToUse, for
* given
* @param includedSample
* covariate will be stored in @param covariate, and column names will be
* stored
* in @colNames
* if covariate file missed some samples, those sample names will be stored in
* @sampleToDrop
* NOTE: for missing values in a covariate, it will drop this covariate out of
* the following anaylysis
* @return number of samples have covariates.
* Example:
* includedSample = [A, B, C] and in covaraite file we have [B, C, C, D]
* then output covariate have 3 rows corresponding to [A, B, C]
* row C filled by the last C in covariate file
* sample D will be in sampleToDrop
*/
int _loadCovariate(const std::string& fn,
const std::vector<std::string>& includedSample,
const std::vector<std::string>& covNameToUse,
DataLoader::HandleMissingCov handleMissingCov,
SimpleMatrix* covariate, std::vector<std::string>* colNames,
std::set<std::string>* sampleToDrop) {
// load covariate
SimpleMatrix mat;
int ret = extractCovariate(fn, includedSample, covNameToUse, handleMissingCov,
&mat, sampleToDrop);
if (ret < 0) {
return -1;
}
// create covariate sample index
// const int nr = mat.nrow();
const int nc = mat.ncol();
std::map<std::string, int> covIndex;
makeMap(mat.getRowName(), &covIndex);
int idx = 0;
for (size_t i = 0; i < includedSample.size(); ++i) {
if (covIndex.find(includedSample[i]) == covIndex.end()) {
sampleToDrop->insert(includedSample[i]);
continue;
}
const int match = covIndex[includedSample[i]];
covariate->resize(idx + 1, nc);
for (int j = 0; j < mat.ncol(); ++j) {
(*covariate)[idx][j] = mat[match][j];
// skip row label, as MathMatrix class does not have row label
}
++idx;
}
// set col label
for (int i = 0; i < mat.ncol(); ++i) {
// (*covariate).SetColumnLabel(i, mat.getColName()[i].c_str());
(*covariate).setColName(i, mat.getColName()[i]);
}
return 0;
} // end _loadCovariate
int main(int argc, char** argv) {
double genotype[] = {0, 2, -9, 2, 2, 2, 2, -9, 1,
2, 2, 2, 2, 1, 1, -9, -9, 0};
SimpleMatrix m; // marker by sample matrix
m.resize(3, 6);
int offset = 0;
for (int i = 0; i < m.nrow(); ++i) {
for (int j = 0; j < m.ncol(); ++j) {
m[i][j] = genotype[offset];
++offset;
}
}
std::vector<std::string> iid;
char buf[128];
for (int i = 0; i < m.ncol(); ++i) {
sprintf(buf, "%d", i + 1);
iid.push_back(buf);
}
std::vector<double> pheno;
pheno.push_back(-9);
pheno.push_back(-9);
pheno.push_back(2);
pheno.push_back(-9);
pheno.push_back(2);
pheno.push_back(2);
PlinkOutputFile pout("testPlinkOutputFile.output");
pout.writeFAM(iid, iid, pheno);
pout.writeBIM("1", "snp1", 0, 1, "G", "A");
pout.writeBIM("1", "snp2", 0, 2, "1", "2");
pout.writeBIM("1", "snp3", 0, 3, "A", "C");
pout.writeBED(&m, m.ncol(), m.nrow());
fprintf(stderr, "Done\n");
return 0;
}
int main(int argc, char *argv[])
{
// check loading all genotypes to memory
PlinkInputFile f1 ("testPlinkInputFile");
SimpleMatrix m;
f1.readIntoMatrix(&m);
for (int i = 0; i < m.nrow(); i++ ) {
for (int j = 0; j < m.ncol(); j++) {
int geno = (int) m[i][j];
printf("%d ", geno);
}
printf("\n");
}
printMatrix(m, f1);
// loading some people and marker
printf("----------------------------------------------------------------------\n");
m.clear();
f1.readIntoMatrix(&m, NULL, NULL);
std::vector<std::string> personName, markerName;
f1.readIntoMatrix(&m, &personName, &markerName);
printMatrix(m, f1);
// loading some people and marker
printf("----------------------------------------------------------------------\n");
m.clear();
std::vector<std::string> people;
std::vector<std::string> marker;
people.push_back("1");
f1.readIntoMatrix(&m, &people, &marker);
printMatrix(m, f1);
// loading some people and marker
printf("----------------------------------------------------------------------\n");
m.clear();
people.push_back("6");
marker.push_back("snp1");
f1.readIntoMatrix(&m, &people, &marker);
printMatrix(m, f1);
// loading some people and marker
printf("----------------------------------------------------------------------\n");
m.clear();
people.push_back("2");
marker.push_back("snp2");
f1.readIntoMatrix(&m, &people, &marker);
printMatrix(m, f1);
return 0;
}
void toMatrix(const SimpleMatrix& from, Matrix* to) {
const int nr = from.nrow();
const int nc = from.ncol();
to->Dimension(nr, nc);
// copy value
for (int i = 0; i < nr; ++i) {
for (int j = 0; j < nc; ++j) {
(*to)(i, j) = from[i][j];
}
}
// copy col labels
for (int i = 0; i < nc; ++i) {
(*to).SetColumnLabel(i, from.getColName()[i].c_str());
}
}