Closed jfsantos closed 8 years ago
Hi,
Yes, I have noticed the issue today.
One must call
U = sc_unchunk{U}
in order to collapse the first two dimensions (time withing chunk and chunk index) into one time dimension.
The fact that data has only 4 elements may be a mistake in bandwidth inequality. Will keep you in touch.
I think the inequality was correct. The issue should be fixed after PR #34. Thank you @jfsantos for your interest.
It works now, thanks! Just a small nitpick: line 35 plots the scattergram over the scalogram, so maybe it would be better to create a new figure before calling imagesc. It's also not clear why you call display_scalogram
twice, the second time after some simple variable manipulation. I guess it's probably there for debugging reasons.
Both remarks are right. Commits 2a77296 and aec2c51 on latest master address them. I have put the two figures as vertical subplots.
That was quick, thanks!
example_handel.m
is currently breaking at line 31:modulation_scale_index
is 8, butdata
has only 4 elements. Also,data
is a 3D array, so the transpose operator is not defined.