matrix n. (pl. matrices 或matrixes) 1.【解剖學】子宮;母體;發(fā)源地,策源地,搖籃;【生物學】襯質(zhì)細胞;間質(zhì);基質(zhì);母質(zhì)。 2. 【礦物】母巖;脈石;【冶金】基體;【地質(zhì)學;地理學】脈石;填質(zhì);雜礦石。 3. 【印刷】字模;型版,紙型;鑄型,陰模。 4.【陣】(矩)陣,方陣;母式;【物理學】間架;【無線電】矩陣變換電路。 5.【染】原色〔紅黃藍白黑五種〕。 the matrix of a nail 【解剖學】指甲床。
Returns the matrix transpose of a given matrix 返回一個給定矩陣的轉(zhuǎn)置矩陣。
Matrix object that is the matrix transpose of the matrix 對象,該對象是此矩陣的轉(zhuǎn)置矩陣。
The results indicate that the serial and parallel optimization for fft , phase compensation and matrix transpose has the most significance for sar imaging performance 分析結(jié)果認為,對sar成像處理的優(yōu)化,主要是對fft 、相位補償運算和矩陣轉(zhuǎn)角運算等3類典型運算的串行優(yōu)化和并行優(yōu)化。
The fast dct algorithm not only can reduce the multiplication number , but also can combine the post - scaling and matrix transpose needed for fast dct with the quantization and scan processes , so as to speed up the whole mpeg encoder . in addition , the table - lookup fast quantization algorithm can further speed up the encoding process 本章提出的快速dct算法不僅減少了所需乘法的次數(shù),而且變換后的后置乘法和矩陣轉(zhuǎn)置過程可與量化和掃描相結(jié)合,進一步加速浙江大學博士學位論文整個mpeg編碼過程。
Also , the relactions between the best block size for matrix transpose and the size and associativity of the processor ' s cache is formulized . for parallel optimization , several programming models available on a numa system , such as lightweight processes ( sproc ) , posix threads , openmp and mpi , are compared , and their speedup and coding complexity are analyzed 對于sar成像處理的并行優(yōu)化,本文對比了在numa架構(gòu)上可用的幾種并行編程模型:輕量級進程、 posix線程、 openmp和mpi ,針對numa架構(gòu)和sar成像處理的特點從加速比、編程復(fù)雜度等多個方面進行了討論。
Chapter 2 analyzes parallel process technology ' s actuality , the requirement of real - time process , and mostly guidelines of parallel process performance . chapter 3 discusses imaging algorithm - - - - - - chirp scaling algorithm theory as well as realization of ideal point target ; and then discuss the scalar of data and operation . chapter 4 discuss the fft and distributed matrix transposing , mostly about ( 1 ) discussed how to realize parallel fft , and evaluate the preformance of parallel fft ; ( 2 ) discuss another step ' s - - - - - - matrix transposing - - - - - - realization can divided into three steps : distributing , renewedly distributing and local transposing of matrix , and then discuss the time of process in detail 第四章分別研究了cs算法中的fft變換和分布式矩陣的轉(zhuǎn)置問題,主要有: ( 1 )對cs算法中運算量最大的步驟fft變換進行了并行性的提取,并對并行fft變換的算法性能進行了評估; ( 2 )分析并研究了cs算法中另一不可或缺的步驟? ?矩陣轉(zhuǎn)置問題,提出矩陣分布、重新分布和局部轉(zhuǎn)置來實現(xiàn)矩陣轉(zhuǎn)置的并行化,并詳細分析了矩陣轉(zhuǎn)置的時間耗費問題。