This commit is contained in:
tqchen
2015-02-09 20:26:39 -08:00
parent 12ee049a74
commit 4a5b9e5f78
14 changed files with 8596 additions and 21 deletions

204
rabit-learn/utils/base64.h Normal file
View File

@@ -0,0 +1,204 @@
#ifndef RABIT_LEARN_UTILS_BASE64_H_
#define RABIT_LEARN_UTILS_BASE64_H_
/*!
* \file base64.h
* \brief data stream support to input and output from/to base64 stream
* base64 is easier to store and pass as text format in mapreduce
* \author Tianqi Chen
*/
#include <cctype>
#include <cstdio>
#include <rabit/io.h>
namespace rabit {
namespace utils {
/*! \brief namespace of base64 decoding and encoding table */
namespace base64 {
const char DecodeTable[] = {
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
62, // '+'
0, 0, 0,
63, // '/'
52, 53, 54, 55, 56, 57, 58, 59, 60, 61, // '0'-'9'
0, 0, 0, 0, 0, 0, 0,
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, // 'A'-'Z'
0, 0, 0, 0, 0, 0,
26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38,
39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, // 'a'-'z'
};
static const char EncodeTable[] =
"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
} // namespace base64
/*! \brief the stream that reads from base64, note we take from file pointers */
class Base64InStream: public IStream {
public:
explicit Base64InStream(FILE *fp) : fp(fp) {
num_prev = 0; tmp_ch = 0;
}
/*!
* \brief initialize the stream position to beginning of next base64 stream
* call this function before actually start read
*/
inline void InitPosition(void) {
// get a charater
do {
tmp_ch = fgetc(fp);
} while (isspace(tmp_ch));
}
/*! \brief whether current position is end of a base64 stream */
inline bool IsEOF(void) const {
return num_prev == 0 && (tmp_ch == EOF || isspace(tmp_ch));
}
virtual size_t Read(void *ptr, size_t size) {
using base64::DecodeTable;
if (size == 0) return 0;
// use tlen to record left size
size_t tlen = size;
unsigned char *cptr = static_cast<unsigned char*>(ptr);
// if anything left, load from previous buffered result
if (num_prev != 0) {
if (num_prev == 2) {
if (tlen >= 2) {
*cptr++ = buf_prev[0];
*cptr++ = buf_prev[1];
tlen -= 2;
num_prev = 0;
} else {
// assert tlen == 1
*cptr++ = buf_prev[0]; --tlen;
buf_prev[0] = buf_prev[1];
num_prev = 1;
}
} else {
// assert num_prev == 1
*cptr++ = buf_prev[0]; --tlen; num_prev = 0;
}
}
if (tlen == 0) return size;
int nvalue;
// note: everything goes with 4 bytes in Base64
// so we process 4 bytes a unit
while (tlen && tmp_ch != EOF && !isspace(tmp_ch)) {
// first byte
nvalue = DecodeTable[tmp_ch] << 18;
{
// second byte
Check((tmp_ch = fgetc(fp), tmp_ch != EOF && !isspace(tmp_ch)),
"invalid base64 format");
nvalue |= DecodeTable[tmp_ch] << 12;
*cptr++ = (nvalue >> 16) & 0xFF; --tlen;
}
{
// third byte
Check((tmp_ch = fgetc(fp), tmp_ch != EOF && !isspace(tmp_ch)),
"invalid base64 format");
// handle termination
if (tmp_ch == '=') {
Check((tmp_ch = fgetc(fp), tmp_ch == '='), "invalid base64 format");
Check((tmp_ch = fgetc(fp), tmp_ch == EOF || isspace(tmp_ch)),
"invalid base64 format");
break;
}
nvalue |= DecodeTable[tmp_ch] << 6;
if (tlen) {
*cptr++ = (nvalue >> 8) & 0xFF; --tlen;
} else {
buf_prev[num_prev++] = (nvalue >> 8) & 0xFF;
}
}
{
// fourth byte
Check((tmp_ch = fgetc(fp), tmp_ch != EOF && !isspace(tmp_ch)),
"invalid base64 format");
if (tmp_ch == '=') {
Check((tmp_ch = fgetc(fp), tmp_ch == EOF || isspace(tmp_ch)),
"invalid base64 format");
break;
}
nvalue |= DecodeTable[tmp_ch];
if (tlen) {
*cptr++ = nvalue & 0xFF; --tlen;
} else {
buf_prev[num_prev ++] = nvalue & 0xFF;
}
}
// get next char
tmp_ch = fgetc(fp);
}
if (kStrictCheck) {
Check(tlen == 0, "Base64InStream: read incomplete");
}
return size - tlen;
}
virtual void Write(const void *ptr, size_t size) {
utils::Error("Base64InStream do not support write");
}
private:
FILE *fp;
int tmp_ch;
int num_prev;
unsigned char buf_prev[2];
// whether we need to do strict check
static const bool kStrictCheck = false;
};
/*! \brief the stream that write to base64, note we take from file pointers */
class Base64OutStream: public IStream {
public:
explicit Base64OutStream(FILE *fp) : fp(fp) {
buf_top = 0;
}
virtual void Write(const void *ptr, size_t size) {
using base64::EncodeTable;
size_t tlen = size;
const unsigned char *cptr = static_cast<const unsigned char*>(ptr);
while (tlen) {
while (buf_top < 3 && tlen != 0) {
buf[++buf_top] = *cptr++; --tlen;
}
if (buf_top == 3) {
// flush 4 bytes out
fputc(EncodeTable[buf[1] >> 2], fp);
fputc(EncodeTable[((buf[1] << 4) | (buf[2] >> 4)) & 0x3F], fp);
fputc(EncodeTable[((buf[2] << 2) | (buf[3] >> 6)) & 0x3F], fp);
fputc(EncodeTable[buf[3] & 0x3F], fp);
buf_top = 0;
}
}
}
virtual size_t Read(void *ptr, size_t size) {
Error("Base64OutStream do not support read");
return 0;
}
/*!
* \brief finish writing of all current base64 stream, do some post processing
* \param endch charater to put to end of stream, if it is EOF, then nothing will be done
*/
inline void Finish(char endch = EOF) {
using base64::EncodeTable;
if (buf_top == 1) {
fputc(EncodeTable[buf[1] >> 2], fp);
fputc(EncodeTable[(buf[1] << 4) & 0x3F], fp);
fputc('=', fp);
fputc('=', fp);
}
if (buf_top == 2) {
fputc(EncodeTable[buf[1] >> 2], fp);
fputc(EncodeTable[((buf[1] << 4) | (buf[2] >> 4)) & 0x3F], fp);
fputc(EncodeTable[(buf[2] << 2) & 0x3F], fp);
fputc('=', fp);
}
buf_top = 0;
if (endch != EOF) fputc(endch, fp);
}
private:
FILE *fp;
int buf_top;
unsigned char buf[4];
};
} // namespace utils
} // namespace rabit
#endif // RABIT_LEARN_UTILS_BASE64_H_

143
rabit-learn/utils/data.h Normal file
View File

@@ -0,0 +1,143 @@
/*!
* Copyright (c) 2015 by Contributors
* \file data.h
* \brief simple data structure that could be used by model
*
* \author Tianqi Chen
*/
#ifndef RABIT_LEARN_DATA_H_
#define RABIT_LEARN_DATA_H_
#include <vector>
#include <cstdlib>
#include <cstdio>
#include <cstring>
#include <limits>
#include <cmath>
#include <rabit.h>
namespace rabit {
// typedef index type
typedef unsigned index_t;
/*! \brief sparse matrix, CSR format */
struct SparseMat {
// sparse matrix entry
struct Entry {
// feature index
index_t findex;
// feature value
float fvalue;
};
// sparse vector
struct Vector {
const Entry *data;
index_t length;
inline const Entry &operator[](size_t i) const {
return data[i];
}
};
inline Vector operator[](size_t i) const {
Vector v;
v.data = &data[0] + row_ptr[i];
v.length = static_cast<index_t>(row_ptr[i + 1]-row_ptr[i]);
return v;
}
// load data from LibSVM format
inline void Load(const char *fname) {
FILE *fi;
if (!strcmp(fname, "stdin")) {
fi = stdin;
} else {
if (strchr(fname, '%') != NULL) {
char s_tmp[256];
snprintf(s_tmp, sizeof(s_tmp), fname, rabit::GetRank());
fi = utils::FopenCheck(s_tmp, "r");
} else {
fi = utils::FopenCheck(fname, "r");
}
}
row_ptr.clear();
row_ptr.push_back(0);
data.clear();
feat_dim = 0;
float label; bool init = true;
char tmp[1024];
while (fscanf(fi, "%s", tmp) == 1) {
Entry e;
unsigned long fidx;
if (sscanf(tmp, "%lu:%f", &fidx, &e.fvalue) == 2) {
e.findex = static_cast<index_t>(fidx);
data.push_back(e);
feat_dim = std::max(fidx, feat_dim);
} else {
if (!init) {
labels.push_back(label);
row_ptr.push_back(data.size());
}
utils::Check(sscanf(tmp, "%f", &label) == 1, "invalid LibSVM format");
init = false;
}
}
// last row
labels.push_back(label);
row_ptr.push_back(data.size());
feat_dim += 1;
utils::Check(feat_dim < std::numeric_limits<index_t>::max(),
"feature dimension exceed limit of index_t"\
"consider change the index_t to unsigned long");
// close the filed
if (fi != stdin) fclose(fi);
}
inline size_t NumRow(void) const {
return row_ptr.size() - 1;
}
// maximum feature dimension
size_t feat_dim;
std::vector<size_t> row_ptr;
std::vector<Entry> data;
std::vector<float> labels;
};
// dense matrix
struct Matrix {
inline void Init(size_t nrow, size_t ncol, float v = 0.0f) {
this->nrow = nrow;
this->ncol = ncol;
data.resize(nrow * ncol);
std::fill(data.begin(), data.end(), v);
}
inline float *operator[](size_t i) {
return &data[0] + i * ncol;
}
inline const float *operator[](size_t i) const {
return &data[0] + i * ncol;
}
inline void Print(const char *fname) {
FILE *fo;
if (!strcmp(fname, "stdout")) {
fo = stdout;
} else {
fo = utils::FopenCheck(fname, "w");
}
for (size_t i = 0; i < data.size(); ++i) {
fprintf(fo, "%g", data[i]);
if ((i+1) % ncol == 0) {
fprintf(fo, "\n");
} else {
fprintf(fo, " ");
}
}
// close the filed
if (fo != stdout) fclose(fo);
}
// number of data
size_t nrow, ncol;
std::vector<float> data;
};
/*!\brief computes a random number modulo the value */
inline int Random(int value) {
return rand() % value;
}
} // namespace rabit
#endif // RABIT_LEARN_DATA_H_

40
rabit-learn/utils/io.h Normal file
View File

@@ -0,0 +1,40 @@
#ifndef RABIT_LEARN_UTILS_IO_H_
#define RABIT_LEARN_UTILS_IO_H_
/*!
* \file io.h
* \brief additional stream interface
* \author Tianqi Chen
*/
namespace rabit {
namespace utils {
/*! \brief implementation of file i/o stream */
class FileStream : public ISeekStream {
public:
explicit FileStream(FILE *fp) : fp(fp) {}
explicit FileStream(void) {
this->fp = NULL;
}
virtual size_t Read(void *ptr, size_t size) {
return std::fread(ptr, size, 1, fp);
}
virtual void Write(const void *ptr, size_t size) {
std::fwrite(ptr, size, 1, fp);
}
virtual void Seek(size_t pos) {
std::fseek(fp, static_cast<long>(pos), SEEK_SET);
}
virtual size_t Tell(void) {
return std::ftell(fp);
}
inline void Close(void) {
if (fp != NULL){
std::fclose(fp); fp = NULL;
}
}
private:
FILE *fp;
};
} // namespace utils
} // namespace rabit
#endif // RABIT_LEARN_UTILS_IO_H_