bnl at bitcave.in-berlin.de (Bernhard Luehr) writes:
>BIOF-MAK at finou.oulu.fi schrieb:
>> We are interested in methods for solving following problems:
>> - to calculate pulse frequency as fuction of time
>I am interestet in the results of your requests to find an algorith for
>pulse frequency analysis.
>I have the problem to scan acoustic patterns of
>luscinia megarhynchos (nightingale). About this case an FFT (fast fourier
>transformation) algorithm is needed.
>I think that the analytic solution must be very similiar to yours problem.
>Perhaps you would answer me, if you have had success.
>If anybody else have a comment or an pretty algorith, please contact me.
There's a whole raft load of work done in endocrine hormone pulse analysis
which may be of interest - the raw material is very noisy being the
results of radioimmune assay and sampling frequency is limited due to
practical problems like exsanguination of the subject !!
I have a package which implements the two most popular algorithms used
for that problem - cluster and pulsar in a neat pc clone piece of
shareware. Email me for details...
|Dr. Ross Lazarus | Insert Standard disclaimer |
|Lecturer in Geriatrics | here |
|Dept.of Community Medicine. | and a suitably amusing |
|Sydney University, Sydney, Australia. | witticism or quote |
|ACSnet --- rml at extro.ucc.su.OZ.AU +=============================+
|SnailMail 29 Francis St. Bondi NSW 2026 Australia |