1
subroutine parsur(iopt,ipar,idim,mu,u,mv,v,f,s,nuest,nvest,
2
* nu,tu,nv,tv,c,fp,wrk,lwrk,iwrk,kwrk,ier)
3
c given the set of ordered points f(i,j) in the idim-dimensional space,
4
c corresponding to grid values (u(i),v(j)) ,i=1,...,mu ; j=1,...,mv,
5
c parsur determines a smooth approximating spline surface s(u,v) , i.e.
7
c ... u(1) <= u <= u(mu) ; v(1) <= v <= v(mv)
9
c with sl(u,v), l=1,2,...,idim bicubic spline functions with common
10
c knots tu(i),i=1,...,nu in the u-variable and tv(j),j=1,...,nv in the
12
c in addition, these splines will be periodic in the variable u if
13
c ipar(1) = 1 and periodic in the variable v if ipar(2) = 1.
14
c if iopt=-1, parsur determines the least-squares bicubic spline
15
c surface according to a given set of knots.
16
c if iopt>=0, the number of knots of s(u,v) and their position
17
c is chosen automatically by the routine. the smoothness of s(u,v) is
18
c achieved by minimalizing the discontinuity jumps of the derivatives
19
c of the splines at the knots. the amount of smoothness of s(u,v) is
20
c determined by the condition that
21
c fp=sumi=1,mu(sumj=1,mv(dist(f(i,j)-s(u(i),v(j)))**2))<=s,
22
c with s a given non-negative constant.
23
c the fit s(u,v) is given in its b-spline representation and can be
24
c evaluated by means of routine surev.
27
c call parsur(iopt,ipar,idim,mu,u,mv,v,f,s,nuest,nvest,nu,tu,
28
c * nv,tv,c,fp,wrk,lwrk,iwrk,kwrk,ier)
31
c iopt : integer flag. unchanged on exit.
32
c on entry iopt must specify whether a least-squares surface
33
c (iopt=-1) or a smoothing surface (iopt=0 or 1)must be
35
c if iopt=0 the routine will start with the initial set of
36
c knots needed for determining the least-squares polynomial
38
c if iopt=1 the routine will continue with the set of knots
39
c found at the last call of the routine.
40
c attention: a call with iopt=1 must always be immediately
41
c preceded by another call with iopt = 1 or iopt = 0.
42
c ipar : integer array of dimension 2. unchanged on exit.
43
c on entry ipar(1) must specify whether (ipar(1)=1) or not
44
c (ipar(1)=0) the splines must be periodic in the variable u.
45
c on entry ipar(2) must specify whether (ipar(2)=1) or not
46
c (ipar(2)=0) the splines must be periodic in the variable v.
47
c idim : integer. on entry idim must specify the dimension of the
48
c surface. 1 <= idim <= 3. unchanged on exit.
49
c mu : integer. on entry mu must specify the number of grid points
50
c along the u-axis. unchanged on exit.
51
c mu >= mumin where mumin=4-2*ipar(1)
52
c u : real array of dimension at least (mu). before entry, u(i)
53
c must be set to the u-co-ordinate of the i-th grid point
54
c along the u-axis, for i=1,2,...,mu. these values must be
55
c supplied in strictly ascending order. unchanged on exit.
56
c mv : integer. on entry mv must specify the number of grid points
57
c along the v-axis. unchanged on exit.
58
c mv >= mvmin where mvmin=4-2*ipar(2)
59
c v : real array of dimension at least (mv). before entry, v(j)
60
c must be set to the v-co-ordinate of the j-th grid point
61
c along the v-axis, for j=1,2,...,mv. these values must be
62
c supplied in strictly ascending order. unchanged on exit.
63
c f : real array of dimension at least (mu*mv*idim).
64
c before entry, f(mu*mv*(l-1)+mv*(i-1)+j) must be set to the
65
c l-th co-ordinate of the data point corresponding to the
66
c the grid point (u(i),v(j)) for l=1,...,idim ,i=1,...,mu
67
c and j=1,...,mv. unchanged on exit.
68
c if ipar(1)=1 it is expected that f(mu*mv*(l-1)+mv*(mu-1)+j)
69
c = f(mu*mv*(l-1)+j), l=1,...,idim ; j=1,...,mv
70
c if ipar(2)=1 it is expected that f(mu*mv*(l-1)+mv*(i-1)+mv)
71
c = f(mu*mv*(l-1)+mv*(i-1)+1), l=1,...,idim ; i=1,...,mu
72
c s : real. on entry (if iopt>=0) s must specify the smoothing
73
c factor. s >=0. unchanged on exit.
74
c for advice on the choice of s see further comments
75
c nuest : integer. unchanged on exit.
76
c nvest : integer. unchanged on exit.
77
c on entry, nuest and nvest must specify an upper bound for the
78
c number of knots required in the u- and v-directions respect.
79
c these numbers will also determine the storage space needed by
80
c the routine. nuest >= 8, nvest >= 8.
81
c in most practical situation nuest = mu/2, nvest=mv/2, will
82
c be sufficient. always large enough are nuest=mu+4+2*ipar(1),
83
c nvest = mv+4+2*ipar(2), the number of knots needed for
84
c interpolation (s=0). see also further comments.
86
c unless ier=10 (in case iopt>=0), nu will contain the total
87
c number of knots with respect to the u-variable, of the spline
88
c surface returned. if the computation mode iopt=1 is used,
89
c the value of nu should be left unchanged between subsequent
90
c calls. in case iopt=-1, the value of nu should be specified
92
c tu : real array of dimension at least (nuest).
93
c on succesful exit, this array will contain the knots of the
94
c splines with respect to the u-variable, i.e. the position of
95
c the interior knots tu(5),...,tu(nu-4) as well as the position
96
c of the additional knots tu(1),...,tu(4) and tu(nu-3),...,
97
c tu(nu) needed for the b-spline representation.
98
c if the computation mode iopt=1 is used,the values of tu(1)
99
c ...,tu(nu) should be left unchanged between subsequent calls.
100
c if the computation mode iopt=-1 is used, the values tu(5),
101
c ...tu(nu-4) must be supplied by the user, before entry.
102
c see also the restrictions (ier=10).
104
c unless ier=10 (in case iopt>=0), nv will contain the total
105
c number of knots with respect to the v-variable, of the spline
106
c surface returned. if the computation mode iopt=1 is used,
107
c the value of nv should be left unchanged between subsequent
108
c calls. in case iopt=-1, the value of nv should be specified
110
c tv : real array of dimension at least (nvest).
111
c on succesful exit, this array will contain the knots of the
112
c splines with respect to the v-variable, i.e. the position of
113
c the interior knots tv(5),...,tv(nv-4) as well as the position
114
c of the additional knots tv(1),...,tv(4) and tv(nv-3),...,
115
c tv(nv) needed for the b-spline representation.
116
c if the computation mode iopt=1 is used,the values of tv(1)
117
c ...,tv(nv) should be left unchanged between subsequent calls.
118
c if the computation mode iopt=-1 is used, the values tv(5),
119
c ...tv(nv-4) must be supplied by the user, before entry.
120
c see also the restrictions (ier=10).
121
c c : real array of dimension at least (nuest-4)*(nvest-4)*idim.
122
c on succesful exit, c contains the coefficients of the spline
123
c approximation s(u,v)
124
c fp : real. unless ier=10, fp contains the sum of squared
125
c residuals of the spline surface returned.
126
c wrk : real array of dimension (lwrk). used as workspace.
127
c if the computation mode iopt=1 is used the values of
128
c wrk(1),...,wrk(4) should be left unchanged between subsequent
130
c lwrk : integer. on entry lwrk must specify the actual dimension of
131
c the array wrk as declared in the calling (sub)program.
132
c lwrk must not be too small.
133
c lwrk >= 4+nuest*(mv*idim+11+4*ipar(1))+nvest*(11+4*ipar(2))+
134
c 4*(mu+mv)+q*idim where q is the larger of mv and nuest.
135
c iwrk : integer array of dimension (kwrk). used as workspace.
136
c if the computation mode iopt=1 is used the values of
137
c iwrk(1),.,iwrk(3) should be left unchanged between subsequent
139
c kwrk : integer. on entry kwrk must specify the actual dimension of
140
c the array iwrk as declared in the calling (sub)program.
141
c kwrk >= 3+mu+mv+nuest+nvest.
142
c ier : integer. unless the routine detects an error, ier contains a
143
c non-positive value on exit, i.e.
144
c ier=0 : normal return. the surface returned has a residual sum of
145
c squares fp such that abs(fp-s)/s <= tol with tol a relat-
146
c ive tolerance set to 0.001 by the program.
147
c ier=-1 : normal return. the spline surface returned is an
148
c interpolating surface (fp=0).
149
c ier=-2 : normal return. the surface returned is the least-squares
150
c polynomial surface. in this extreme case fp gives the
151
c upper bound for the smoothing factor s.
152
c ier=1 : error. the required storage space exceeds the available
153
c storage space, as specified by the parameters nuest and
155
c probably causes : nuest or nvest too small. if these param-
156
c eters are already large, it may also indicate that s is
158
c the approximation returned is the least-squares surface
159
c according to the current set of knots. the parameter fp
160
c gives the corresponding sum of squared residuals (fp>s).
161
c ier=2 : error. a theoretically impossible result was found during
162
c the iteration proces for finding a smoothing surface with
163
c fp = s. probably causes : s too small.
164
c there is an approximation returned but the corresponding
165
c sum of squared residuals does not satisfy the condition
167
c ier=3 : error. the maximal number of iterations maxit (set to 20
168
c by the program) allowed for finding a smoothing surface
169
c with fp=s has been reached. probably causes : s too small
170
c there is an approximation returned but the corresponding
171
c sum of squared residuals does not satisfy the condition
173
c ier=10 : error. on entry, the input data are controlled on validity
174
c the following restrictions must be satisfied.
175
c -1<=iopt<=1, 0<=ipar(1)<=1, 0<=ipar(2)<=1, 1 <=idim<=3
176
c mu >= 4-2*ipar(1),mv >= 4-2*ipar(2), nuest >=8, nvest >= 8,
177
c kwrk>=3+mu+mv+nuest+nvest,
178
c lwrk >= 4+nuest*(mv*idim+11+4*ipar(1))+nvest*(11+4*ipar(2))
179
c +4*(mu+mv)+max(nuest,mv)*idim
180
c u(i-1)<u(i),i=2,..,mu, v(i-1)<v(i),i=2,...,mv
181
c if iopt=-1: 8<=nu<=min(nuest,mu+4+2*ipar(1))
182
c u(1)<tu(5)<tu(6)<...<tu(nu-4)<u(mu)
183
c 8<=nv<=min(nvest,mv+4+2*ipar(2))
184
c v(1)<tv(5)<tv(6)<...<tv(nv-4)<v(mv)
185
c the schoenberg-whitney conditions, i.e. there must
186
c be subset of grid co-ordinates uu(p) and vv(q) such
187
c that tu(p) < uu(p) < tu(p+4) ,p=1,...,nu-4
188
c tv(q) < vv(q) < tv(q+4) ,q=1,...,nv-4
189
c (see fpchec or fpchep)
191
c if s=0: nuest>=mu+4+2*ipar(1)
192
c nvest>=mv+4+2*ipar(2)
193
c if one of these conditions is found to be violated,control
194
c is immediately repassed to the calling program. in that
195
c case there is no approximation returned.
198
c by means of the parameter s, the user can control the tradeoff
199
c between closeness of fit and smoothness of fit of the approximation.
200
c if s is too large, the surface will be too smooth and signal will be
201
c lost ; if s is too small the surface will pick up too much noise. in
202
c the extreme cases the program will return an interpolating surface
203
c if s=0 and the constrained least-squares polynomial surface if s is
204
c very large. between these extremes, a properly chosen s will result
205
c in a good compromise between closeness of fit and smoothness of fit.
206
c to decide whether an approximation, corresponding to a certain s is
207
c satisfactory the user is highly recommended to inspect the fits
209
c recommended values for s depend on the accuracy of the data values.
210
c if the user has an idea of the statistical errors on the data, he
211
c can also find a proper estimate for s. for, by assuming that, if he
212
c specifies the right s, parsur will return a surface s(u,v) which
213
c exactly reproduces the surface underlying the data he can evaluate
214
c the sum(dist(f(i,j)-s(u(i),v(j)))**2) to find a good estimate for s.
215
c for example, if he knows that the statistical errors on his f(i,j)-
216
c values is not greater than 0.1, he may expect that a good s should
217
c have a value not larger than mu*mv*(0.1)**2.
218
c if nothing is known about the statistical error in f(i,j), s must
219
c be determined by trial and error, taking account of the comments
220
c above. the best is then to start with a very large value of s (to
221
c determine the le-sq polynomial surface and the corresponding upper
222
c bound fp0 for s) and then to progressively decrease the value of s
223
c ( say by a factor 10 in the beginning, i.e. s=fp0/10,fp0/100,...
224
c and more carefully as the approximation shows more detail) to
225
c obtain closer fits.
226
c to economize the search for a good s-value the program provides with
227
c different modes of computation. at the first call of the routine, or
228
c whenever he wants to restart with the initial set of knots the user
230
c if iopt = 1 the program will continue with the knots found at
231
c the last call of the routine. this will save a lot of computation
232
c time if parsur is called repeatedly for different values of s.
233
c the number of knots of the surface returned and their location will
234
c depend on the value of s and on the complexity of the shape of the
235
c surface underlying the data. if the computation mode iopt = 1
236
c is used, the knots returned may also depend on the s-values at
237
c previous calls (if these were smaller). therefore, if after a number
238
c of trials with different s-values and iopt=1,the user can finally
239
c accept a fit as satisfactory, it may be worthwhile for him to call
240
c parsur once more with the chosen value for s but now with iopt=0.
241
c indeed, parsur may then return an approximation of the same quality
242
c of fit but with fewer knots and therefore better if data reduction
243
c is also an important objective for the user.
244
c the number of knots may also depend on the upper bounds nuest and
245
c nvest. indeed, if at a certain stage in parsur the number of knots
246
c in one direction (say nu) has reached the value of its upper bound
247
c (nuest), then from that moment on all subsequent knots are added
248
c in the other (v) direction. this may indicate that the value of
249
c nuest is too small. on the other hand, it gives the user the option
250
c of limiting the number of knots the routine locates in any direction
251
c for example, by setting nuest=8 (the lowest allowable value for
252
c nuest), the user can indicate that he wants an approximation with
253
c splines which are simple cubic polynomials in the variable u.
255
c other subroutines required:
256
c fppasu,fpchec,fpchep,fpknot,fprati,fpgrpa,fptrnp,fpback,
257
c fpbacp,fpbspl,fptrpe,fpdisc,fpgivs,fprota
261
c dept. computer science, k.u. leuven
262
c celestijnenlaan 200a, b-3001 heverlee, belgium.
263
c e-mail : Paul.Dierckx@cs.kuleuven.ac.be
265
c latest update : march 1989
268
c ..scalar arguments..
270
integer iopt,idim,mu,mv,nuest,nvest,nu,nv,lwrk,kwrk,ier
271
c ..array arguments..
272
real*8 u(mu),v(mv),f(mu*mv*idim),tu(nuest),tv(nvest),
273
* c((nuest-4)*(nvest-4)*idim),wrk(lwrk)
274
integer ipar(2),iwrk(kwrk)
276
real*8 tol,ub,ue,vb,ve,peru,perv
277
integer i,j,jwrk,kndu,kndv,knru,knrv,kwest,l1,l2,l3,l4,
278
* lfpu,lfpv,lwest,lww,maxit,nc,mf,mumin,mvmin
279
c ..function references..
281
c ..subroutine references..
282
c fppasu,fpchec,fpchep
284
c we set up the parameters tol and maxit.
287
c before starting computations a data check is made. if the input data
288
c are invalid, control is immediately repassed to the calling program.
290
if(iopt.lt.(-1) .or. iopt.gt.1) go to 200
291
if(ipar(1).lt.0 .or. ipar(1).gt.1) go to 200
292
if(ipar(2).lt.0 .or. ipar(2).gt.1) go to 200
293
if(idim.le.0 .or. idim.gt.3) go to 200
295
if(mu.lt.mumin .or. nuest.lt.8) go to 200
297
if(mv.lt.mvmin .or. nvest.lt.8) go to 200
299
nc = (nuest-4)*(nvest-4)
300
lwest = 4+nuest*(mv*idim+11+4*ipar(1))+nvest*(11+4*ipar(2))+
301
* 4*(mu+mv)+max0(nuest,mv)*idim
302
kwest = 3+mu+mv+nuest+nvest
303
if(lwrk.lt.lwest .or. kwrk.lt.kwest) go to 200
305
if(u(i-1).ge.u(i)) go to 200
308
if(v(i-1).ge.v(i)) go to 200
310
if(iopt.ge.0) go to 100
311
if(nu.lt.8 .or. nu.gt.nuest) go to 200
314
if (ipar(1).ne.0) go to 40
321
call fpchec(u,mu,tu,nu,3,ier)
322
if(ier.ne.0) go to 200
339
call fpchep(u,mu,tu,nu,3,ier)
340
if(ier.ne.0) go to 200
341
60 if(nv.lt.8 .or. nv.gt.nvest) go to 200
344
if (ipar(2).ne.0) go to 80
351
call fpchec(v,mv,tv,nv,3,ier)
352
if(ier.ne.0) go to 200
369
call fpchep(v,mv,tv,nv,3,ier)
370
if (ier.eq.0) go to 150
372
100 if(s.lt.0.) go to 200
373
if(s.eq.0. .and. (nuest.lt.(mu+4+2*ipar(1)) .or.
374
* nvest.lt.(mv+4+2*ipar(2))) )go to 200
376
c we partition the working space and determine the spline approximation
380
jwrk = lwrk-4-nuest-nvest
385
call fppasu(iopt,ipar,idim,u,mu,v,mv,f,mf,s,nuest,nvest,
386
* tol,maxit,nc,nu,tu,nv,tv,c,fp,wrk(1),wrk(2),wrk(3),wrk(4),
387
* wrk(lfpu),wrk(lfpv),iwrk(1),iwrk(2),iwrk(3),iwrk(knru),
388
* iwrk(knrv),iwrk(kndu),iwrk(kndv),wrk(lww),jwrk,ier)