void ifaGroup::detectTwoTier()
{
int mlen = maxAbilities;
if (!twotier || mlen < 3) return;
std::vector<int> orthogonal;
if (mlen >= 3) {
Eigen::Map<Eigen::MatrixXd> Ecov(cov, mlen, mlen);
Eigen::Matrix<Eigen::DenseIndex, Eigen::Dynamic, 1> numCov((Ecov.array() != 0.0).matrix().colwise().count());
std::vector<int> candidate;
for (int fx=0; fx < numCov.rows(); ++fx) {
if (numCov(fx) == 1) candidate.push_back(fx);
}
if (candidate.size() > 1) {
std::vector<bool> mask(numItems());
for (int cx=candidate.size() - 1; cx >= 0; --cx) {
std::vector<bool> loading(numItems());
for (int ix=0; ix < numItems(); ++ix) {
loading[ix] = param[ix * paramRows + candidate[cx]] != 0;
}
std::vector<bool> overlap(loading.size());
std::transform(loading.begin(), loading.end(),
mask.begin(), overlap.begin(),
std::logical_and<bool>());
if (std::find(overlap.begin(), overlap.end(), true) == overlap.end()) {
std::transform(loading.begin(), loading.end(),
mask.begin(), mask.begin(),
std::logical_or<bool>());
orthogonal.push_back(candidate[cx]);
}
}
}
std::reverse(orthogonal.begin(), orthogonal.end());
}
if (orthogonal.size() == 1) orthogonal.clear();
if (orthogonal.size() && orthogonal[0] != mlen - int(orthogonal.size())) {
Rf_error("Independent factors must be given after dense factors");
}
numSpecific = orthogonal.size();
if (numSpecific) {
Sgroup.assign(numItems(), 0);
for (int ix=0; ix < numItems(); ix++) {
for (int dx=orthogonal[0]; dx < maxAbilities; ++dx) {
if (param[ix * paramRows + dx] != 0) {
Sgroup[ix] = dx - orthogonal[0];
continue;
}
}
}
}
}
void normalToStdVector(omxMatrix *cov, omxMatrix *mean, omxMatrix *slope, omxMatrix *thr,
int numOrdinal, std::vector< omxThresholdColumn > &ti,
Eigen::Ref<Eigen::VectorXd> out)
{
// order of elements: (c.f. lav_model_wls, lavaan 0.6-2)
// 1. thresholds + means (interleaved)
// 2. slopes (if any, columnwise per exo)
// 3. variances (continuous indicators only)
// 4. covariances; not correlations (lower triangle)
EigenMatrixAdaptor Ecov(cov);
if (numOrdinal == 0) {
int dx = 0;
if (mean) {
EigenVectorAdaptor Emean(mean);
for (int rx=0; rx < cov->cols; ++rx) {
out[dx++] = Emean(rx);
}
}
if (slope) {
EigenMatrixAdaptor Eslope(slope);
for (int cx=0; cx < Eslope.cols(); ++cx) {
for (int rx=0; rx < Eslope.rows(); ++rx) {
out[dx++] = Eslope(rx,cx);
}
}
}
for (int cx=0; cx < cov->cols; ++cx) {
out[dx++] = Ecov(cx,cx);
}
for (int cx=0; cx < cov->cols-1; ++cx) {
for (int rx=cx+1; rx < cov->rows; ++rx) {
out[dx++] = Ecov(rx,cx);
}
}
return;
}
if (!mean) Rf_error("ordinal indicators and no mean vector");
EigenVectorAdaptor Emean(mean);
EigenMatrixAdaptor Eth(thr);
Eigen::VectorXd sdTmp(1.0/Ecov.diagonal().array().sqrt());
Eigen::DiagonalMatrix<double, Eigen::Dynamic> sd(Emean.size());
sd.setIdentity();
int dx = 0;
for (auto &th : ti) {
for (int t1=0; t1 < th.numThresholds; ++t1) {
double sd1 = sdTmp[th.dColumn];
out[dx++] = (Eth(t1, th.column) - Emean[th.dColumn]) * sd1;
sd.diagonal()[th.dColumn] = sd1;
}
if (!th.numThresholds) {
out[dx++] = Emean[th.dColumn];
}
}
if (slope) {
EigenMatrixAdaptor Eslope(slope);
for (int cx=0; cx < Eslope.cols(); ++cx) {
for (int rx=0; rx < Eslope.rows(); ++rx) {
out[dx++] = Eslope(rx,cx);
}
}
}
Eigen::MatrixXd stdCov(sd * Ecov * sd);
for (int cx=0; cx < cov->cols; ++cx) {
if (ti[cx].numThresholds) continue;
out[dx++] = stdCov(cx,cx);
}
for (int cx=0; cx < cov->cols-1; ++cx) {
for (int rx=cx+1; rx < cov->rows; ++rx) {
out[dx++] = stdCov(rx,cx);
}
}
}