You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.

dlatm2.f 9.1 kB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315
  1. *> \brief \b DLATM2
  2. *
  3. * =========== DOCUMENTATION ===========
  4. *
  5. * Online html documentation available at
  6. * http://www.netlib.org/lapack/explore-html/
  7. *
  8. * Definition:
  9. * ===========
  10. *
  11. * DOUBLE PRECISION FUNCTION DLATM2( M, N, I, J, KL, KU, IDIST,
  12. * ISEED, D, IGRADE, DL, DR, IPVTNG, IWORK, SPARSE )
  13. *
  14. * .. Scalar Arguments ..
  15. *
  16. * INTEGER I, IDIST, IGRADE, IPVTNG, J, KL, KU, M, N
  17. * DOUBLE PRECISION SPARSE
  18. * ..
  19. *
  20. * .. Array Arguments ..
  21. *
  22. * INTEGER ISEED( 4 ), IWORK( * )
  23. * DOUBLE PRECISION D( * ), DL( * ), DR( * )
  24. * ..
  25. *
  26. *
  27. *> \par Purpose:
  28. * =============
  29. *>
  30. *> \verbatim
  31. *>
  32. *> DLATM2 returns the (I,J) entry of a random matrix of dimension
  33. *> (M, N) described by the other parameters. It is called by the
  34. *> DLATMR routine in order to build random test matrices. No error
  35. *> checking on parameters is done, because this routine is called in
  36. *> a tight loop by DLATMR which has already checked the parameters.
  37. *>
  38. *> Use of DLATM2 differs from SLATM3 in the order in which the random
  39. *> number generator is called to fill in random matrix entries.
  40. *> With DLATM2, the generator is called to fill in the pivoted matrix
  41. *> columnwise. With DLATM3, the generator is called to fill in the
  42. *> matrix columnwise, after which it is pivoted. Thus, DLATM3 can
  43. *> be used to construct random matrices which differ only in their
  44. *> order of rows and/or columns. DLATM2 is used to construct band
  45. *> matrices while avoiding calling the random number generator for
  46. *> entries outside the band (and therefore generating random numbers
  47. *>
  48. *> The matrix whose (I,J) entry is returned is constructed as
  49. *> follows (this routine only computes one entry):
  50. *>
  51. *> If I is outside (1..M) or J is outside (1..N), return zero
  52. *> (this is convenient for generating matrices in band format).
  53. *>
  54. *> Generate a matrix A with random entries of distribution IDIST.
  55. *>
  56. *> Set the diagonal to D.
  57. *>
  58. *> Grade the matrix, if desired, from the left (by DL) and/or
  59. *> from the right (by DR or DL) as specified by IGRADE.
  60. *>
  61. *> Permute, if desired, the rows and/or columns as specified by
  62. *> IPVTNG and IWORK.
  63. *>
  64. *> Band the matrix to have lower bandwidth KL and upper
  65. *> bandwidth KU.
  66. *>
  67. *> Set random entries to zero as specified by SPARSE.
  68. *> \endverbatim
  69. *
  70. * Arguments:
  71. * ==========
  72. *
  73. *> \param[in] M
  74. *> \verbatim
  75. *> M is INTEGER
  76. *> Number of rows of matrix. Not modified.
  77. *> \endverbatim
  78. *>
  79. *> \param[in] N
  80. *> \verbatim
  81. *> N is INTEGER
  82. *> Number of columns of matrix. Not modified.
  83. *> \endverbatim
  84. *>
  85. *> \param[in] I
  86. *> \verbatim
  87. *> I is INTEGER
  88. *> Row of entry to be returned. Not modified.
  89. *> \endverbatim
  90. *>
  91. *> \param[in] J
  92. *> \verbatim
  93. *> J is INTEGER
  94. *> Column of entry to be returned. Not modified.
  95. *> \endverbatim
  96. *>
  97. *> \param[in] KL
  98. *> \verbatim
  99. *> KL is INTEGER
  100. *> Lower bandwidth. Not modified.
  101. *> \endverbatim
  102. *>
  103. *> \param[in] KU
  104. *> \verbatim
  105. *> KU is INTEGER
  106. *> Upper bandwidth. Not modified.
  107. *> \endverbatim
  108. *>
  109. *> \param[in] IDIST
  110. *> \verbatim
  111. *> IDIST is INTEGER
  112. *> On entry, IDIST specifies the type of distribution to be
  113. *> used to generate a random matrix .
  114. *> 1 => UNIFORM( 0, 1 )
  115. *> 2 => UNIFORM( -1, 1 )
  116. *> 3 => NORMAL( 0, 1 )
  117. *> Not modified.
  118. *> \endverbatim
  119. *>
  120. *> \param[in,out] ISEED
  121. *> \verbatim
  122. *> ISEED is INTEGER array of dimension ( 4 )
  123. *> Seed for random number generator.
  124. *> Changed on exit.
  125. *> \endverbatim
  126. *>
  127. *> \param[in] D
  128. *> \verbatim
  129. *> D is DOUBLE PRECISION array of dimension ( MIN( I , J ) )
  130. *> Diagonal entries of matrix. Not modified.
  131. *> \endverbatim
  132. *>
  133. *> \param[in] IGRADE
  134. *> \verbatim
  135. *> IGRADE is INTEGER
  136. *> Specifies grading of matrix as follows:
  137. *> 0 => no grading
  138. *> 1 => matrix premultiplied by diag( DL )
  139. *> 2 => matrix postmultiplied by diag( DR )
  140. *> 3 => matrix premultiplied by diag( DL ) and
  141. *> postmultiplied by diag( DR )
  142. *> 4 => matrix premultiplied by diag( DL ) and
  143. *> postmultiplied by inv( diag( DL ) )
  144. *> 5 => matrix premultiplied by diag( DL ) and
  145. *> postmultiplied by diag( DL )
  146. *> Not modified.
  147. *> \endverbatim
  148. *>
  149. *> \param[in] DL
  150. *> \verbatim
  151. *> DL is DOUBLE PRECISION array ( I or J, as appropriate )
  152. *> Left scale factors for grading matrix. Not modified.
  153. *> \endverbatim
  154. *>
  155. *> \param[in] DR
  156. *> \verbatim
  157. *> DR is DOUBLE PRECISION array ( I or J, as appropriate )
  158. *> Right scale factors for grading matrix. Not modified.
  159. *> \endverbatim
  160. *>
  161. *> \param[in] IPVTNG
  162. *> \verbatim
  163. *> IPVTNG is INTEGER
  164. *> On entry specifies pivoting permutations as follows:
  165. *> 0 => none.
  166. *> 1 => row pivoting.
  167. *> 2 => column pivoting.
  168. *> 3 => full pivoting, i.e., on both sides.
  169. *> Not modified.
  170. *> \endverbatim
  171. *>
  172. *> \param[out] IWORK
  173. *> \verbatim
  174. *> IWORK is INTEGER array ( I or J, as appropriate )
  175. *> This array specifies the permutation used. The
  176. *> row (or column) in position K was originally in
  177. *> position IWORK( K ).
  178. *> This differs from IWORK for DLATM3. Not modified.
  179. *> \endverbatim
  180. *>
  181. *> \param[in] SPARSE
  182. *> \verbatim
  183. *> SPARSE is DOUBLE PRECISION between 0. and 1.
  184. *> On entry specifies the sparsity of the matrix
  185. *> if sparse matrix is to be generated.
  186. *> SPARSE should lie between 0 and 1.
  187. *> A uniform ( 0, 1 ) random number x is generated and
  188. *> compared to SPARSE; if x is larger the matrix entry
  189. *> is unchanged and if x is smaller the entry is set
  190. *> to zero. Thus on the average a fraction SPARSE of the
  191. *> entries will be set to zero.
  192. *> Not modified.
  193. *> \endverbatim
  194. *
  195. * Authors:
  196. * ========
  197. *
  198. *> \author Univ. of Tennessee
  199. *> \author Univ. of California Berkeley
  200. *> \author Univ. of Colorado Denver
  201. *> \author NAG Ltd.
  202. *
  203. *> \date June 2016
  204. *
  205. *> \ingroup double_matgen
  206. *
  207. * =====================================================================
  208. DOUBLE PRECISION FUNCTION DLATM2( M, N, I, J, KL, KU, IDIST,
  209. $ ISEED, D, IGRADE, DL, DR, IPVTNG, IWORK, SPARSE )
  210. *
  211. * -- LAPACK auxiliary routine (version 3.7.0) --
  212. * -- LAPACK is a software package provided by Univ. of Tennessee, --
  213. * -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
  214. * June 2016
  215. *
  216. * .. Scalar Arguments ..
  217. *
  218. INTEGER I, IDIST, IGRADE, IPVTNG, J, KL, KU, M, N
  219. DOUBLE PRECISION SPARSE
  220. * ..
  221. *
  222. * .. Array Arguments ..
  223. *
  224. INTEGER ISEED( 4 ), IWORK( * )
  225. DOUBLE PRECISION D( * ), DL( * ), DR( * )
  226. * ..
  227. *
  228. * =====================================================================
  229. *
  230. * .. Parameters ..
  231. *
  232. DOUBLE PRECISION ZERO
  233. PARAMETER ( ZERO = 0.0D0 )
  234. * ..
  235. *
  236. * .. Local Scalars ..
  237. *
  238. INTEGER ISUB, JSUB
  239. DOUBLE PRECISION TEMP
  240. * ..
  241. *
  242. * .. External Functions ..
  243. *
  244. DOUBLE PRECISION DLARAN, DLARND
  245. EXTERNAL DLARAN, DLARND
  246. * ..
  247. *
  248. *-----------------------------------------------------------------------
  249. *
  250. * .. Executable Statements ..
  251. *
  252. *
  253. * Check for I and J in range
  254. *
  255. IF( I.LT.1 .OR. I.GT.M .OR. J.LT.1 .OR. J.GT.N ) THEN
  256. DLATM2 = ZERO
  257. RETURN
  258. END IF
  259. *
  260. * Check for banding
  261. *
  262. IF( J.GT.I+KU .OR. J.LT.I-KL ) THEN
  263. DLATM2 = ZERO
  264. RETURN
  265. END IF
  266. *
  267. * Check for sparsity
  268. *
  269. IF( SPARSE.GT.ZERO ) THEN
  270. IF( DLARAN( ISEED ).LT.SPARSE ) THEN
  271. DLATM2 = ZERO
  272. RETURN
  273. END IF
  274. END IF
  275. *
  276. * Compute subscripts depending on IPVTNG
  277. *
  278. IF( IPVTNG.EQ.0 ) THEN
  279. ISUB = I
  280. JSUB = J
  281. ELSE IF( IPVTNG.EQ.1 ) THEN
  282. ISUB = IWORK( I )
  283. JSUB = J
  284. ELSE IF( IPVTNG.EQ.2 ) THEN
  285. ISUB = I
  286. JSUB = IWORK( J )
  287. ELSE IF( IPVTNG.EQ.3 ) THEN
  288. ISUB = IWORK( I )
  289. JSUB = IWORK( J )
  290. END IF
  291. *
  292. * Compute entry and grade it according to IGRADE
  293. *
  294. IF( ISUB.EQ.JSUB ) THEN
  295. TEMP = D( ISUB )
  296. ELSE
  297. TEMP = DLARND( IDIST, ISEED )
  298. END IF
  299. IF( IGRADE.EQ.1 ) THEN
  300. TEMP = TEMP*DL( ISUB )
  301. ELSE IF( IGRADE.EQ.2 ) THEN
  302. TEMP = TEMP*DR( JSUB )
  303. ELSE IF( IGRADE.EQ.3 ) THEN
  304. TEMP = TEMP*DL( ISUB )*DR( JSUB )
  305. ELSE IF( IGRADE.EQ.4 .AND. ISUB.NE.JSUB ) THEN
  306. TEMP = TEMP*DL( ISUB ) / DL( JSUB )
  307. ELSE IF( IGRADE.EQ.5 ) THEN
  308. TEMP = TEMP*DL( ISUB )*DL( JSUB )
  309. END IF
  310. DLATM2 = TEMP
  311. RETURN
  312. *
  313. * End of DLATM2
  314. *
  315. END