- 04 Oct, 2015 1 commit
-
-
thomas.forbriger authored
-
- 23 Sep, 2015 3 commits
-
-
thomas.forbriger authored
compile.sh might return an error und thus terminate make
-
thomas.forbriger authored
compile.sh is designed to be used by other software systems too. For this reason it should be installed along with binary executables.
-
thomas.forbriger authored
gremlin.tex used pepi.sty and pepi.bst for which copyright is not fully resolved. Replaced these entries with a standard bibtex style just to make gremlin.tex compile outside my environment.
-
- 22 Sep, 2015 1 commit
-
-
thomas.forbriger authored
-
- 21 Sep, 2015 11 commits
-
-
thomas.forbriger authored
-
thomas.forbriger authored
-
thomas.forbriger authored
-
thomas.forbriger authored
Point out potentially misleading results of the application of the convolution function to time series being understood as cut-out of infinite functions.
-
thomas.forbriger authored
The test cases provided here demonstrate that the convolution function produces a (possibly unwanted) step response if one of the signals ends with a finite value. This is theoretically correct, however might be unintended in practical application.
-
thomas.forbriger authored
-
thomas.forbriger authored
-
thomas.forbriger authored
-
thomas.forbriger authored
-
thomas.forbriger authored
-
thomas.forbriger authored
Provide test plot T1Z.ps which focusses on the erroneous coda in the time domain filter response when applying the taper prior to the filter.
-
- 20 Sep, 2015 4 commits
-
-
thomas.forbriger authored
Application of taper after filter removes the faulty signal, because taper has length of original signal only, not of result of convolution operation.
-
thomas.forbriger authored
This operation does not produce the expected results. Leave this to be fixed later.
-
thomas.forbriger authored
swapping order of filter an taper application makes faulty signal vanish
-
thomas.forbriger authored
Convolve with delta pulse to delay by given time interval. To be used as a test case for libtsxx convolve.h
-
- 19 Sep, 2015 1 commit
-
-
thomas.forbriger authored
-
- 18 Sep, 2015 4 commits
-
-
thomas.forbriger authored
-
thomas.forbriger authored
-
thomas.forbriger authored
-
thomas.forbriger authored
The result of the libtsxx convolution operation appears unsatisfactory. In order to test the library function I require a binary program to be applied to test series. Since all other operations in cross.cc are identical it appears reasonable to provide a convolution option in cross.cc Also added: scaling with sampling interval
-
- 31 Jul, 2015 14 commits
-
-
thomas.forbriger authored
program is not experimental
-
thomas.forbriger authored
and make tidesffx a standard target
-
thomas.forbriger authored
-
thomas.forbriger authored
-
thomas.forbriger authored
missed to add two new files to the previous commit
-
thomas.forbriger authored
-
thomas.forbriger authored
-
thomas.forbriger authored
-
thomas.forbriger authored
-
thomas.forbriger authored
-
thomas.forbriger authored
-
thomas.forbriger authored
Some issues of improper Fortran code are fixed. There are a many more remaining (implicite type conversion with value truncation, unused labels, unused variables, use of unititialized variable). I do not correct them here since practically all of them affect the original 3rd-party code.
-
thomas.forbriger authored
gfortran does not appreciate Hollerith code
-
thomas.forbriger authored
Move programs for computation of tidal signals to subdirectory tides. Adjust Makefile and list of directories with installable programs. Adjust README and COPYING files also with respect to Seitosh layout.
-
- 30 Jul, 2015 1 commit
-
-
thomas.forbriger authored
-