Unfortunately I've had the same issue and ended up exporting the data manually using MATLAB's dlmwrite. I am not using miremotion, though. You could try writing a function to do the same thing or try to figure out the structure mirexport(...,'Raw') uses... -- João Felipe Santos On Fri, Feb 8, 2013 at 11:52 AM, Maureen Aw <maureen_510@xxxxxxxxxxx> wrote: > Hi João > > > Thanks for your reply. But apparently I got 1303 columns of raw data, > which I have got no idea what they represent. Any ideas? > > Maureen > > On 08-Feb-2013, at 11:57 PM, João Felipe Santos wrote: > > Hi Maureen, > > I've just started to use mirexport for my data, but the manual tells there > is a 'Raw' option that you can add to the parameters which would skip the > use of mirstat to compute statistics of the features. Here's what's on the > manual: > > "If the keyword mirexport(..., ‘Raw’) is added, what is exported is the > complete, raw, feature > data instead. Feature data are stored in columns, and arrays shorter than > others are completed > with NaN values in order to obtain a matrix." > > Hope it helps! > > -- > João Felipe Santos > > > On Fri, Feb 8, 2013 at 10:10 AM, Maureen Aw <maureen_510@xxxxxxxxxxx>wrote: > >> Hi, >> >> I have a simple question. Is there an efficient way of extracting the >> values from the miremotion plot? I understand that there is a way to attain >> the co-ordinates from the plot, but I have more than 300 songs that I am >> examining, so I hope to be able to export the values all at once, if >> possible. I tried mirexport, but instead I got the mean, standard >> deviation, slope, period freq, period amplitude and period entropy, in >> which period freq, period amplitude results are NaN. In addition, I do not >> know how to utilize the values from mirexport to attain the co-ordinates I >> see from the plot. >> >> Any suggestions? Thanks in advance! >> >> Maureen >> > > >