Bugs with Ziel - result dissociation corrected
minor
Writing docs.
Version 0.6.0
Reintegrating uDFT branch into trunk
uFFT rewritten for use of open arrays and functions instead of procedures
For InsertFrom, variant with OpenArray added
forprint of vectors, open array are used
uMatrix: open arrays instead of TVector used in operators
Convolution works for signal length = 1
Add an lmath project_group for easier first manual usage
Target created in lpg (attachment was empty) and included into the library.
uFindChebyshevCoeffs.pas added; some error codes added and used in lmDSP/lmdsp.lpk; in some places open array parameters used instead of TVector.
1. uFFT and uDFT moved from lmRegression to lmDSP
Working uDFT, corrected uFFT, uCompVecUtils, Euler constant, predefined variables Pi and Euler in uEval, new Demo program for Fourier transforms.
Hi! Many thanks for your work! Little problem is that I am user of Lazarus release version myself and cannot try your work before including in the package, though in principle I find it potentially very useful. I think, for now I leave this ticket open, and include your work in LMath library immediately after project groups are included into release of Lazarus. Many thanks once more!
Add an lmath project_group for easier first manual usage
slightly optimized
Lb, Ub calling interface and small optimizations. Never called functions commented out.
Optimizations of VectorHelper
Histo demo program compiled and tested.
Khi2 works if calculated theoretical value in distribution is 0.
Most of BGI compiles and works at Windows 64
Adding Test program to UDFT folder
Branching for DFT
removed uDFT
uDFT added to lmRegression
LMath 0.5.1
Removed custom complilation options in lmGenMath
New names of the directories, everywhere removed high optimization options, missing files in Demo included
Final version of uCobyla and uTrsTlp
Correct finalization of TrsTlp and Cobyla
z in utrstlp.pas is properly initialized which fixes bug in uCobyla
Hi! nv_icons.lrs together with original .png files are included in the distributed file lmath.zip. Being a sole developer I did not see much sense in including them into svn. But as I see now, many users prefer to check out svn rather than download distribution files, I added the icons to svn. At the next svn update you will have them. Best regards, and thanks for reports and interesting discussion at the forum.
Adding icons to svn
Thanks! lmath.lpk and all it's packages compile fine now. However I have the following error while trying to compile lmcomponents.lpk: Hint: (11030) Start of reading config file C:\PRG\Lazarus\Fixes20x32x\fpc\bin\i386-win32\fpc.cfg Hint: (11031) End of reading config file C:\PRG\Lazarus\Fixes20x32x\fpc\bin\i386-win32\fpc.cfg Free Pascal Compiler version 3.2.1-r45984 [2020/08/02] for i386 Copyright (c) 1993-2020 by Florian Klaempfl and others (1002) Target OS: Win32 for i386 (3104) Compiling lmcomponents.pas...
lmath.lpk created which imports all LMath packages. Now it is possible for a program to import lmath.lpk instead of importing every single package.
Error in FindSigma corrected
Hi! Many thanks for your report. Indeed, seems that I forgot to commit these two files, now they are in the repo. Now they are committed. But generally, I recommend to download zip files from here rather then from the repository, unless you are interested in the latest development. I use Trunk for ongoing development, hence it may be untested and even not working. Best regards, Viatcheslav
Can not compile LMath 0.5 with 32bit FPC 3.2.1 and Lazarus 2.0.11 on Win10 x64
Hi! Many thanks for your report. Indeed, seems that I forgot to commit these two files, now they are in the repo. But generally, I recommend to download zip files from here rather then from the repository, unless you are interested in the latest development. I use Trunk for ongoing development, hence it may be untested and even not working. Best regards, Viatcheslav
Adding uVecFunc and uvecmatprn.pas
Can not compile LMath 0.5 with 32bit FPC 3.2.1 and Lazarus 2.0.11 on Win10 x64
lide editing with history browsing
Creating tag corresponding to Ver 0.5
St and En replaced with Lb and Ub
uVectorHelper used in ConstrFit/fitlogexp.pas
long prefixes can be used
MaxNum constant for 64-bit arch defined with conditional compliation
FastDraw takes TRealPointVector instead of TPoints
declarations of classes added
distExtract* procedures added
demo programs for spline and distribs added
error in HypoExponentialDistribution2 corrected
types.inc: parameters V1, V2 changed to Ref, Val in TComparator and TIntComparator.