< Back to IRCAM Forum

Pattern p-markov-analyze

Hi all!

I can’t get p-markov-analyze to print the transition-table, is it possible that :print? is broken?

Thanks!

Hi

Please can you state at least where can we find this function since it is not natively implemented in OM ?
This will be of some help.
Best
K

Sorry, I wasn’t clear with the subject (and misspelled it in addition…): it’s in the patterns library GitHub - openmusic-project/Patterns: An adaption to OpenMusic of Rick Taube's "Pattern Streams"
In the tutorial patch it says

:print {true | :table | :pattern | false}

If the value is true (the default) then two results are
printed, a transition table and a pattern description. The
transition table that is printer details the distribution of
inputs and outputs in the data. The pattern description that
is printed will instantiate that table if it is
evalutated. If :print is just :table or :pattern then only
that form of the analysis is printed. If :print is false then
no analysis is printed.

but I can’t get the transition-table printed.
Thanks!

I just got it working in om# right away, I have to re-check on the other machine with om…

On the other machine with om 6.15 it isn’t working, just connected (1 2 1 2 3) and t.

Hi

Better use the latest version of OM (6.18)
Beside thst i will leave the author of the library (Anders) to reply on this issue. you’re right, first the default value of :print is nil and not 't. if true is used no print.

Best
K

Thanks for pointing this out!

It should be fixed now in the current git sources. If you want an immediate cure you can replace the code for markov-analyze in sources/pattern-classes.lisp with the attached.

markov-analyze.lisp (4.5 KB)

-anders

Thank you very much!

Another question concerning p-markov-analyze: is there a method to get access to the data of the transition table to slightly manipulate them?

Thanks!

Hi pfchph,

if you supply :pattern to the value of the :print? keyword, p-markov-analyze prints out a lisp-expression which, when evaluated, returns a markov-pattern. You can modify the data and weights freely in this expression, before evaluating it.

Another way, which will work for all pattern classes, could be to ask the pattern to return it’s data, using the function ompa::pattern-data.

Thank you very much!