Method insertCorpus

src/python/handler/py_LDA.cpp:31–109

}

std::vector<size_t> LDAModelObject::insertCorpus(PyObject* corpusObj, PyObject* transform)
{
	vector<size_t> ret;
	if (!corpusObj || corpusObj == Py_None) return ret;
	if (!PyObject_TypeCheck(corpusObj, py::Type<CorpusObject>)) throw py::ValueError{ "`corpus` must be an instance of `tomotopy.utils.Corpus`" };
	auto corpus = (CorpusObject*)corpusObj;
	bool insert_into_empty = inst->updateVocab(corpus->getVocabDict().getRaw());
	if (corpus->isIndependent())
	{
		for (auto& rdoc : corpus->docs)
		{
			tomoto::RawDoc doc;
			doc.rawStr = rdoc.rawStr;
			doc.weight = rdoc.weight;
			doc.docUid = rdoc.docUid;

			for (size_t i = 0; i < rdoc.words.size(); ++i)
			{
				if (rdoc.words[i] == tomoto::non_vocab_id) continue;

				if (insert_into_empty) doc.words.emplace_back(rdoc.words[i]);
				else doc.words.emplace_back(corpus->getVocabDict().mapToNewDict(rdoc.words[i], inst->getVocabDict()));

				if (!doc.rawStr.empty())
				{
					doc.origWordPos.emplace_back(rdoc.origWordPos[i]);
					doc.origWordLen.emplace_back(rdoc.origWordLen[i]);
				}
			}

			if (doc.words.empty())
			{
				fprintf(stderr, "Adding empty document was ignored.\n");
				continue;
			}

			if (!doc.rawStr.empty()) char2Byte(doc.rawStr, doc.origWordPos, doc.origWordLen);
			doc.misc = convertMisc(transformMisc(rdoc.misc, transform));
			ret.emplace_back(inst->addDoc(doc));
		}
	}
	else
	{
		for (Py_ssize_t i = 0; i < corpus->len(); ++i)
		{
			auto& rdoc = (tomoto::DocumentBase&)*corpus->getDoc(i);

			tomoto::RawDoc doc;
			doc.rawStr = rdoc.rawStr;
			doc.weight = rdoc.weight;
			doc.docUid = rdoc.docUid;

			for (size_t i = 0; i < rdoc.words.size(); ++i)
			{
				if (rdoc.words[i] == tomoto::non_vocab_id) continue;

				doc.words.emplace_back(corpus->getVocabDict().mapToNewDict(rdoc.words[i], inst->getVocabDict()));