Re: Convert double to octave_value for assignment to column vector

[babelproofreader]

> If I were you, I would declare i_slopes and friends as a ColumnVector
> (as opposed to a OCTAVE_LOCAL_BUFFER'd double*). 

This works, and I had thought of doing this myself. The thing is, the
function rep_median_compile.cc is just an interim function for me to
learn/develop my coding skills. Ultimately the function I want to write will
be an adaptive, moving window application of the repeated median fit, with
various length look back periods determined by a second input column to the
function. This being the case, I think I really need to learn/be shown the
proper syntax for the use of OCTAVE_LOCAL_BUFFER, as in the final function
it will not be possible to predefine the length of i_slopes etc. The problem
as I now see it is how does one pass the contents of an OCTAVE_LOCAL_BUFFER
to a feval median call?
-- 
View this message in context: 
http://octave.1599824.n4.nabble.com/Convert-double-to-octave-value-for-assignment-to-column-vector-tp2239451p2246640.html
Sent from the Octave - General mailing list archive at Nabble.com.