Frank Cwitkowitz

this should fix issue #1 and cover the intentions of pull request #3

I think either solution would be fine, however, from a user-pointing perspective the experiment name is just a string, so it seems one should be able to specify any arbitrary...

Here is some information on my use-case. I am using GuitarSet, which has note and pitch annotations in JAMS. When evaluating MF0E performance, I would like to use the pitch...

@rabitt / @craffel How does the following sound for a solution? Scrap what I have so far and replace it with a warning similar to that in the melody resampling...

> This seems like a reasonable solution to me at a high level. Great. I can get started on this then.

@rabitt - The script you linked would certainly work, however it requires that the sampling rate and hop length are known. In my case these values are not explicitly defined....

@rabitt - how does this look? I can add test cases once we are happy with the changes.

Good suggestion @rabitt I made that change and added a few other small changes necessary to get the test cases working. Here a a few final considerations before completing the...

@rabitt - I recently came back to this and ran into an error where sometimes the final (original) observation, while still occurring within the specified duration, extends in time beyond...

I think the best solution here is simply to add an extra frame in all cases, even if this means the nearest frame on the uniform time grid for the...