SCM

SCM Repository

[matrix] Diff of /pkg/TODO
ViewVC logotype

Diff of /pkg/TODO

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 510, Fri Feb 4 23:36:17 2005 UTC revision 956, Fri Sep 30 17:28:00 2005 UTC
# Line 1  Line 1 
1    - Sparse matrix methods can now be based on the CHOLMOD package.  We
2       will need to migrate from the current code to CHOLMOD-based code
3       using #ifdef USE_CHOLMOD.  Some of the things to be done
4    
5       - Move documentation from subdirectories of src to inst/doc
6       - Write utilities to create a cholmod_sparse pointer from a
7       dgCMatrix or lgCMatrix (or zgCMatrix) object without copying and
8       allocating.
9       - Start adding simple S4 methods (rcond, %*%, +, cbind, t).
10    
11  - Report the problem in the Linux ldexp manual page.  The second and  - Report the problem in the Linux ldexp manual page.  The second and
12    third calls in the Synopsis should be to ldexpf and ldexpl.    third calls in the Synopsis should be to ldexpf and ldexpl.
13    
14  - [,] indexing  - [,] indexing: for sparse "works", but not yet for negative indices!
15    
16    - group generics: "Arith" (partly done),
17      but also "Compare", "Math" etc;
18      see ?Math  and the examples in ?SetGeneric
19    
20    - methods for rbind and cbind where they make sense
21      (Kurt is particularly interested in dgCMatrix ones, for pkg 'arules')
22      ****> UNFEASIBLE as long cbind is function(..., *) !!!! <*******
23        --> follow John's proposition  cbind(x, ...) via Generic cbind2(x,y)?
24    
25    - consider moving alloc3Darray from ./src/Mutils.c to
26      $(RSRC)/src/base/array.c
27    
28    -------
29    
30    - provide methods for "dspMatrix" and "dppMatrix"!
31    
32    - implement (more) methods for supporting "packed" (symmetric / triangular)
33      matrices; particularly something like pack() and unpack()  [to/from our
34      classes from/to "numeric"] --- have already man/unpack.Rd but no method yet!
35    
36      (have some dtr* <-> dtp*)
37    
38    - implement diagonal Matrix class  "ddiMatrix" etc
39      using constructor function Diagonal() or Diag().
40    
41    BUG: rcond() of a singular dpoMatrix gives a LaPack error instead of just 0:
42      MM <- crossprod(M <- Matrix(c(1:4,9:6), 2,4)) ; rcond(MM)
43      ##> Error in rcond(MM) : Lapack routine dpotrf returned error code 4
44    
 - group generics: "Arith", but also "Ops" and "Math";  see  
   ?Math.data.frame  and the examples in  
   ?SetGeneric  
45    
46  - check to see if the .onLoad function to require the methods package  ---
   (in the AllClass.R file) is needed  
47    
48  - Should the uplo and diag slots continue to be stored as character?  - combine the C functions for multiplication by special forms and
49    An alternative is to use a factor as in the enum values for the    solution wrt special forms by using a 'right' argument and a
50    cblas.    'classed' argument.
51       [done with dgeMatrix_matrix_mm();  not yet for other classes;
52        and for _crossprod()]
53    
54  - Organization of the source code files - right now they are organized  - add more comprehensive examples / tests for Schur decomposition
   according to class (e.g. dgeMatrix.R, dgeMatrix.h, dgeMatrix.c).  Is  
   there a better way?  
55    
56  - Fix the calculation of the Dim slot for the crossprod method for  - arithmetic for sparse matrices:
57    dgCMatrix objects (too tired to do that now).               <sparseMatrix>  o  <same-dim-sparseMatrix>
58      should return a sparse matrix  for at least "+" and "*" , also %%,
59      and "/" and "%/%" at least when the RHS is non-zero a scalar.
60      Challenge: nice implementation (``common non-0''; but Tsparse* is not uniq).
61    
62  - bCrosstab(): do we really want the diagonal "V:V" crosstabs?  ---
63    
64  - src/Metis/ : some Makefile needs fixing, as changing src/Metis/*.c  - Create a Harwell-Boeing version of the matrix mm and the response
65               does not lead to recompilation.    vector y in inst/external and remove them from the data directory.
66      Modify any examples that use them and modify the Comparisons vignette.
67    
68  - man/Matrix.Rd :  has example with dimnames, but we just drop them!  - "Math2" , "Math", "Arith":
69                  MM thinks dimnames should be supported (but then ...)     keep triangular and symmetric Matrices when appropriate:
70       particularly desirable for  "Math2": round(), signif()
71    
72  - data/ : names 'mm' and even more 'y' are  ``too short''.  - "d" <-> "l" coercion for all "[TCR]" sparse matrices is really trivial:
73          If we really want to keep them, don't use "LazyData"    "d" -> "l" : drops the 'x' slot
74          (such that one needs  data(*) explicitly);    "l" -> "d" : construct an 'x' slot of all '1'
75          But MM would rather want something like  ex.mm and ex.y    We currently have many of these conversions explicitly, e.g.
76       setAs("dsTMatrix", "lsTMatrix",
77          function(from) new("lsTMatrix", i = from@i, j = from@j, uplo = from@uplo,
78                             Dim = from@Dim, Dimnames = from@Dimnames))
79      but I would rather want to automatically construct all these coercion
80      methods at once by a ``method constructor'', i.e.,
81      for all  "dsparse*" -> "lsparse*" and vice versa.
82      How can one do this {in a documented way} ?

Legend:
Removed from v.510  
changed lines
  Added in v.956

root@r-forge.r-project.org
ViewVC Help
Powered by ViewVC 1.0.0  
Thanks to:
Vienna University of Economics and Business Powered By FusionForge