Rbm/source/RnnTextHelper.cpp

/*
 * To change this license header, choose License Headers in Project Properties.
 * To change this template file, choose Tools | Templates
 * and open the template in the editor.
 */

/* 
 * File:   RnnTextHelper.cpp
 * Author: jens
 * 
 * Created on 18. Januar 2022, 19:12
 */

#include "RnnTextHelper.hpp"
#include "matutils.hpp"

using namespace std;

RnnTextHelper::RnnTextHelper()
{
}

RnnTextHelper::~RnnTextHelper()
{
}

int RnnTextHelper::char_is(char c, const char* pTable)
{
	int found = 0;
	while (*pTable)
	{
		if (c == *pTable)
		{
			found = 1;
			break;
		}
		pTable++;
	}
	return found;
}

int RnnTextHelper::ch2idx(char c)
{
	c = toupper(c);

	int index = 0;
	if (isalpha(c))
	{
		index = 1 + (int) (c - 'A');
	}
	if (isdigit(c))
	{
		index = 1 + 26 + (int) (c - '0');
	}
	return index;
}

char RnnTextHelper::idx2ch(int index)
{
	char c = '?';
	if (index == 0)
	{
		c = ' ';
	}
	else if (index >= 1 and index < 27)
	{
		c = (char) (index + 'A' - 1);
	}
	else if (index >= 27 and index < 38)
	{
		c = (char) (index + '0' - 27);
	}
	return c;
}

arma::mat RnnTextHelper::createTraining(const string& filename, size_t seq_len)
{
	const char SPACE = ' ';
	FILE *pFile;

	pFile = fopen(filename.c_str(), "r");

	if (!pFile)
	{
		std::cout << "Could not open " << filename << "!" << std::endl;
		return 0;
	}

	int numTraining = 0;
	int numVisible = seq_len*NUM_CODES;

	arma::mat batch = arma::zeros(numTraining, numVisible);
	arma::mat pattern = arma::zeros(seq_len, NUM_CODES);

	for (int i = 0; i < seq_len; i++)
	{
		pattern.row(i) = Matutils::char2vec(SPACE, NUM_CODES);
	}

	int last_index = ch2idx(SPACE);
	while (!feof(pFile))
	{
		char ch = SPACE;
		int result = fread(&ch, 1, 1, pFile);
		if (result < 0)
		{
			break;
		}

		int index = ch2idx(ch);
		if (index == 0 and last_index == 0)
		{
			continue;
		}
		last_index = index;
		
		if (seq_len > 1)
		{
			pattern = arma::shift(pattern, 1);
		}
		pattern.row(0) = Matutils::char2vec(ch, NUM_CODES);
		arma::mat pattern_vector = pattern.as_row();
		batch.insert_rows(batch.n_rows, pattern_vector);
	}
	fclose(pFile);

	return batch;
}