sz_float.c 13.6 KB
Newer Older
A
Alex Duan 已提交
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 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 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434
/**
 *  @file sz_float.c
 *  @author Sheng Di, Dingwen Tao, Xin Liang, Xiangyu Zou, Tao Lu, Wen Xia, Xuan Wang, Weizhe Zhang
 *  @date Aug, 2016
 *  @brief SZ_Init, Compression and Decompression functions
 *  (C) 2016 by Mathematics and Computer Science (MCS), Argonne National Laboratory.
 *      See COPYRIGHT in top-level directory.
 */


#include <stddef.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <math.h>
#include "sz.h"
#include "CompressElement.h"
#include "DynamicByteArray.h"
#include "TightDataPointStorageF.h"
#include "sz_float.h"
#include "szd_float.h"
#include "utility.h"

# define MIN(_a, _b) (((_a) < (_b)) ? (_a) : (_b))



void computeReqLength_float(double realPrecision, short rangeExpo, int* reqLength, float* medianValue)
{
	short realPrecExpo = getPrecisionReqLength_double(realPrecision);
	*reqLength = 9 + rangeExpo - realPrecExpo + 1; //radExpo-reqExpo == reqMantiLength
	if(*reqLength<9)
		*reqLength = 9;
	if(*reqLength>32)
	{	
		*reqLength = 32;
		*medianValue = 0;
	}			
}


unsigned int optimize_intervals_float_1D(float *oriData, size_t dataLength, double realPrecision)
{	
	size_t i = 0, radiusIndex;
	float pred_value = 0, pred_err;
	size_t *intervals = (size_t*)malloc(confparams_cpr->maxRangeRadius*sizeof(size_t));
	memset(intervals, 0, confparams_cpr->maxRangeRadius*sizeof(size_t));
	size_t totalSampleSize = dataLength/confparams_cpr->sampleDistance;
	for(i=2;i<dataLength;i++)
	{
		if(i%confparams_cpr->sampleDistance==0)
		{
			//pred_value = 2*oriData[i-1] - oriData[i-2];
			pred_value = oriData[i-1];
			pred_err = fabs(pred_value - oriData[i]);
			radiusIndex = (unsigned long)((pred_err/realPrecision+1)/2);
			if(radiusIndex>=confparams_cpr->maxRangeRadius)
				radiusIndex = confparams_cpr->maxRangeRadius - 1;			
			intervals[radiusIndex]++;
		}
	}
	//compute the appropriate number
	size_t targetCount = totalSampleSize*confparams_cpr->predThreshold;
	size_t sum = 0;
	for(i=0;i<confparams_cpr->maxRangeRadius;i++)
	{
		sum += intervals[i];
		if(sum>targetCount)
			break;
	}
	if(i>=confparams_cpr->maxRangeRadius)
		i = confparams_cpr->maxRangeRadius-1;
		
	unsigned int accIntervals = 2*(i+1);
	unsigned int powerOf2 = roundUpToPowerOf2(accIntervals);
	
	if(powerOf2<32)
		powerOf2 = 32;
	
	free(intervals);
	//printf("accIntervals=%d, powerOf2=%d\n", accIntervals, powerOf2);
	return powerOf2;
}

TightDataPointStorageF* SZ_compress_float_1D_MDQ(float *oriData, 
				size_t dataLength, float realPrecision, float valueRangeSize, float medianValue_f)
{
	unsigned int quantization_intervals;
	if(exe_params->optQuantMode==1)
		quantization_intervals = optimize_intervals_float_1D_opt(oriData, dataLength, realPrecision);
	else
		quantization_intervals = exe_params->intvCapacity;
	//updateQuantizationInfo(quantization_intervals);	
	int half_interval = quantization_intervals/2;

    //
	// calc reqlength and need set medianValue to zero. 
	// 
	size_t i;
	int reqLength; // need save bits length for one float .  value ragne 9~32 
	float medianValue = medianValue_f;
	short rangeExpo = getExponent_float(valueRangeSize/2);
	
	computeReqLength_float(realPrecision, rangeExpo, &reqLength, &medianValue);	

	//
	//  malloc all 
	//
	
	// 1 type 
	int* type = (int*) malloc(dataLength*sizeof(int));	
	float* spaceFillingValue = oriData; //
	// 2 lead
	DynamicIntArray *exactLeadNumArray;
	new_DIA(&exactLeadNumArray, DynArrayInitLen);
	// 3 mid
	DynamicByteArray *exactMidByteArray;
	new_DBA(&exactMidByteArray, DynArrayInitLen);
	// 4 residual bit
	DynamicIntArray *resiBitArray;
	new_DIA(&resiBitArray, DynArrayInitLen);
	
	unsigned char preDiffBytes[4];
	intToBytes_bigEndian(preDiffBytes, 0);
	
	// calc save byte length and bit lengths with reqLength
	int reqBytesLength = reqLength/8;
	int resiBitsLength = reqLength%8;
	//float last3CmprsData[3] = {0};

	FloatValueCompressElement *vce = (FloatValueCompressElement*)malloc(sizeof(FloatValueCompressElement));
	LossyCompressionElement *lce = (LossyCompressionElement*)malloc(sizeof(LossyCompressionElement));
				
	//add the first data	
	type[0] = 0;
	compressSingleFloatValue(vce, spaceFillingValue[0], realPrecision, medianValue, reqLength, reqBytesLength, resiBitsLength);
	// set lce
	updateLossyCompElement_Float(vce->curBytes, preDiffBytes, reqBytesLength, resiBitsLength, lce);
	memcpy(preDiffBytes, vce->curBytes, 4);
	// lce to arrays
	addExactData(exactMidByteArray, exactLeadNumArray, resiBitArray, lce);
	//listAdd_float(last3CmprsData, vce->data);
		
	//add the second data
	type[1] = 0;
	compressSingleFloatValue(vce, spaceFillingValue[1], realPrecision, medianValue, reqLength, reqBytesLength, resiBitsLength);
	updateLossyCompElement_Float(vce->curBytes, preDiffBytes, reqBytesLength, resiBitsLength, lce);
	memcpy(preDiffBytes, vce->curBytes, 4);
	addExactData(exactMidByteArray, exactLeadNumArray, resiBitArray, lce);
	//listAdd_float(last3CmprsData, vce->data);

	int state;
	float checkRadius;
	float oriFloat;
	float pred = vce->data;
	float diff;
	checkRadius = (quantization_intervals-1)*realPrecision;
	float double_realpreci = 2*realPrecision;
	float recip_precision = 1/realPrecision;
	
	for(i=2; i < dataLength; i++)
	{	
		oriFloat = spaceFillingValue[i];
		//pred = 2*last3CmprsData[0] - last3CmprsData[1];
		//pred = last3CmprsData[0];
		diff = fabsf(oriFloat - pred);	
		if(diff < checkRadius)
		{
			state = ((int)( diff * recip_precision + 1))>>1;
			if(oriFloat >= pred)
			{
				type[i] = half_interval + state;
				pred = pred + state * double_realpreci;
			}
			else //curData<pred
			{
				type[i] = half_interval - state;
				pred = pred - state * double_realpreci;
			}
				
			//double-check the prediction error in case of machine-epsilon impact	
			if(fabs(oriFloat - pred) > realPrecision)
			{	
				type[i] = 0;				
				compressSingleFloatValue(vce, oriFloat, realPrecision, medianValue, reqLength, reqBytesLength, resiBitsLength);
				updateLossyCompElement_Float(vce->curBytes, preDiffBytes, reqBytesLength, resiBitsLength, lce);
				memcpy(preDiffBytes, vce->curBytes, 4);
				addExactData(exactMidByteArray, exactLeadNumArray, resiBitArray, lce);		
				
				//listAdd_float(last3CmprsData, vce->data);	
				pred = vce->data;				
			}
			else
			{
				//listAdd_float(last3CmprsData, pred);	
			}	
			// go next
			continue;
		}
		
		//unpredictable data processing		
		type[i] = 0;		
		compressSingleFloatValue(vce, oriFloat, realPrecision, medianValue, reqLength, reqBytesLength, resiBitsLength);
		updateLossyCompElement_Float(vce->curBytes, preDiffBytes, reqBytesLength, resiBitsLength, lce);
		memcpy(preDiffBytes, vce->curBytes, 4);
		addExactData(exactMidByteArray, exactLeadNumArray, resiBitArray, lce);

		//listAdd_float(last3CmprsData, vce->data);
		pred = vce->data;	
		
	}//end of for
		
//	char* expSegmentsInBytes;
//	int expSegmentsInBytes_size = convertESCToBytes(esc, &expSegmentsInBytes);
	size_t exactDataNum = exactLeadNumArray->size;
	
	TightDataPointStorageF* tdps = NULL;			
	new_TightDataPointStorageF(&tdps, dataLength, exactDataNum, 
			type, exactMidByteArray->array, exactMidByteArray->size,  
			exactLeadNumArray->array,  
			resiBitArray->array, resiBitArray->size, 
			resiBitsLength,
			realPrecision, medianValue, (char)reqLength, quantization_intervals, 0);

//sdi:Debug
/*	int sum =0;
	for(i=0;i<dataLength;i++)
		if(type[i]==0) sum++;
	printf("opt_quantizations=%d, exactDataNum=%zu, sum=%d\n",quantization_intervals, exactDataNum, sum);
*/	
	//free memory
	free_DIA(exactLeadNumArray);
	free_DIA(resiBitArray);
	free(type);	
	free(vce);
	free(lce);	
	free(exactMidByteArray); //exactMidByteArray->array has been released in free_TightDataPointStorageF(tdps);
	
	return tdps;
}

// compress core algorithm  if success return true else return false
bool SZ_compress_args_float_NoCkRngeNoGzip_1D( unsigned char* newByteData, float *oriData, 
                   size_t dataLength, double realPrecision, size_t *outSize, float valueRangeSize, float medianValue_f)
{		
	// get tdps
	TightDataPointStorageF* tdps = NULL;	
	tdps = SZ_compress_float_1D_MDQ(oriData, dataLength, realPrecision, valueRangeSize, medianValue_f);	
	if(tdps == NULL)
	  return false;

    // serialize 
	if(!convertTDPStoFlatBytes_float(tdps, newByteData, outSize))
	{
		free_TightDataPointStorageF(tdps);
		return false;
	}
	  
	// check compressed size large than original
	if(*outSize > 1 + MetaDataByteLength + exe_params->SZ_SIZE_TYPE + 1 + sizeof(float)*dataLength)
	{
		return false;
	}	
	free_TightDataPointStorageF(tdps);
	return true;
}


void SZ_compress_args_float_withinRange(unsigned char* newByteData, float *oriData, size_t dataLength, size_t *outSize)
{
	TightDataPointStorageF* tdps = (TightDataPointStorageF*) malloc(sizeof(TightDataPointStorageF));
	memset(tdps, 0, sizeof(TightDataPointStorageF));
	tdps->leadNumArray = NULL;
	tdps->residualMidBits = NULL;
	
	tdps->allSameData = 1;
	tdps->dataSeriesLength = dataLength;
	tdps->exactMidBytes = (unsigned char*)malloc(sizeof(unsigned char)*4);
	tdps->isLossless = 0;
	float value = oriData[0];
	floatToBytes(tdps->exactMidBytes, value);
	tdps->exactMidBytes_size = 4;
	
	size_t tmpOutSize;
	//unsigned char *tmpByteData;
	convertTDPStoFlatBytes_float(tdps, newByteData, &tmpOutSize);

	//*newByteData = (unsigned char*)malloc(sizeof(unsigned char)*12); //for floating-point data (1+3+4+4)
	//memcpy(*newByteData, tmpByteData, 12);
	*outSize = tmpOutSize; //8+SZ_SIZE_TYPE; //8==3+1+4(float_size)
	free_TightDataPointStorageF(tdps);	
}

void cost_start();
double cost_end(const char* tag);
void show_rate( int in_len, int out_len);

int SZ_compress_args_float(float *oriData, size_t r1, unsigned char* newByteData, size_t *outSize, sz_params* params)
{
	int status = SZ_SUCCESS;
	size_t dataLength = r1;
	
	//cost_start();
	// check at least elements count  
	if(dataLength <= MIN_NUM_OF_ELEMENTS)
	{
		printf("warning, input elements count=%d less than %d, so need not do compress.\n", (int)dataLength, MIN_NUM_OF_ELEMENTS);
		return SZ_LITTER_ELEMENT;
	}
	
	float valueRangeSize = 0, medianValue = 0;
	float min = 0;

	min = computeRangeSize_float(oriData, dataLength, &valueRangeSize, &medianValue);	

	// calc max
	float max = min+valueRangeSize;
	params->fmin = min;
	params->fmax = max;
	
	// calc precision
	double realPrecision = 0; 
	if(params->errorBoundMode==PSNR)
	{
		params->errorBoundMode = SZ_ABS;
		realPrecision = params->absErrBound = computeABSErrBoundFromPSNR(params->psnr, (double)params->predThreshold, (double)valueRangeSize);
		//printf("realPrecision=%lf\n", realPrecision);
	}
	else if(params->errorBoundMode==NORM) //norm error = sqrt(sum((xi-xi_)^2))
	{
		params->errorBoundMode = SZ_ABS;
		realPrecision = params->absErrBound = computeABSErrBoundFromNORM_ERR(params->normErr, dataLength);
		//printf("realPrecision=%lf\n", realPrecision);				
	}
	else
	{
		realPrecision = getRealPrecision_float(valueRangeSize, params->errorBoundMode, params->absErrBound, params->relBoundRatio, &status);
		params->absErrBound = realPrecision;
	}	

	//cost_end(" sz_pre_calc");

    //
	// do compress 
	//
	if(valueRangeSize <= realPrecision)
	{		
		// special deal with same data
		SZ_compress_args_float_withinRange(newByteData, oriData, dataLength, outSize);
		return SZ_SUCCESS;
	}

	//
	// first compress with sz
	//		
	size_t tmpOutSize = 0;
	unsigned char* tmpByteData  =  newByteData;
	bool twoStage =  params->szMode != SZ_BEST_SPEED;
	if(twoStage)
	{
		tmpByteData = (unsigned char*)malloc(r1*sizeof(float) + 1024);
	}

	// compress core algorithm
	if(!SZ_compress_args_float_NoCkRngeNoGzip_1D(tmpByteData, oriData, r1, realPrecision, &tmpOutSize, valueRangeSize, medianValue))
	{
		*outSize = 0;
		if(twoStage)
			free(tmpByteData);
		return SZ_ALGORITHM_ERR;
	}

    //cost_end(" sz_first_compress");
	//show_rate(r1*sizeof(float), tmpOutSize);
	//
	//  second compress with Call Zstd or Gzip 
	//
	//cost_start();
	if(!twoStage)
	{
		*outSize = tmpOutSize;
	}
	else
	{
		*outSize = sz_lossless_compress(params->losslessCompressor, tmpByteData, tmpOutSize, newByteData);
		free(tmpByteData);	
	}

	//cost_end(" sz_second_compress");
	//show_rate(r1*sizeof(float), *outSize);
	return SZ_SUCCESS;
}

unsigned int optimize_intervals_float_1D_opt(float *oriData, size_t dataLength, double realPrecision)
{	
	size_t i = 0, radiusIndex;
	float pred_value = 0, pred_err;
	size_t *intervals = (size_t*)malloc(confparams_cpr->maxRangeRadius*sizeof(size_t));
	memset(intervals, 0, confparams_cpr->maxRangeRadius*sizeof(size_t));
	size_t totalSampleSize = 0;//dataLength/confparams_cpr->sampleDistance;

	float * data_pos = oriData + 2;
	while(data_pos - oriData < dataLength){
		totalSampleSize++;
		pred_value = data_pos[-1];
		pred_err = fabs(pred_value - *data_pos);
		radiusIndex = (unsigned long)((pred_err/realPrecision+1)/2);
		if(radiusIndex>=confparams_cpr->maxRangeRadius)
			radiusIndex = confparams_cpr->maxRangeRadius - 1;			
		intervals[radiusIndex]++;

		data_pos += confparams_cpr->sampleDistance;
	}
	//compute the appropriate number
	size_t targetCount = totalSampleSize*confparams_cpr->predThreshold;
	size_t sum = 0;
	for(i=0;i<confparams_cpr->maxRangeRadius;i++)
	{
		sum += intervals[i];
		if(sum>targetCount)
			break;
	}
	if(i>=confparams_cpr->maxRangeRadius)
		i = confparams_cpr->maxRangeRadius-1;
		
	unsigned int accIntervals = 2*(i+1);
	unsigned int powerOf2 = roundUpToPowerOf2(accIntervals);
	
	if(powerOf2<32)
		powerOf2 = 32;
	
	free(intervals);
	return powerOf2;
}