log

age author description
Tue, 20 May 2014 13:36:20 +0100 Chris Cannam More informative error output
Tue, 20 May 2014 13:36:02 +0100 Chris Cannam Add read-all
Thu, 15 May 2014 16:22:12 +0100 Chris Cannam Limit white noise to +/-0.5
Tue, 13 May 2014 18:50:27 +0100 Chris Cannam Comment only
Tue, 13 May 2014 16:42:07 +0100 Chris Cannam Fix dirac "window" length (and fix test)
Mon, 12 May 2014 17:46:31 +0100 Chris Cannam Fixes to tests for frequency preservation in resampling
Mon, 12 May 2014 17:46:19 +0100 Chris Cannam Fix delay in decimator (apply it before picking)
Mon, 12 May 2014 11:29:50 +0100 Chris Cannam Need abs diff
Mon, 12 May 2014 11:24:47 +0100 Chris Cannam Test that all convolution methods produce the same output for a longer stream
Mon, 12 May 2014 11:24:24 +0100 Chris Cannam Provide yetidir
Sun, 11 May 2014 16:02:43 +0100 Chris Cannam Some failing resampler tests, matching those currently in qm-dsp
Thu, 01 May 2014 11:36:14 +0100 Chris Cannam Add sweeps and time-varying sinusoids
Wed, 30 Apr 2014 13:32:14 +0100 Chris Cannam Fixes to vector sizing in resampler
Wed, 30 Apr 2014 10:35:13 +0100 Chris Cannam Better error reporting
Sat, 05 Apr 2014 16:02:24 +0100 Chris Cannam Merge
Mon, 03 Mar 2014 17:03:32 +0000 Chris Cannam Comment
Mon, 03 Mar 2014 17:02:06 +0000 Chris Cannam Comment
Tue, 01 Apr 2014 10:35:45 +0100 Chris Cannam Sorted, median, quantiles
Mon, 31 Mar 2014 14:40:47 +0100 Chris Cannam Print (some of) the stack trace when catching exception from test
Mon, 31 Mar 2014 14:31:56 +0100 Chris Cannam Make vector arithmetic functions throw if args are of differing lengths -- definitely makes for a higher probability of getting correct code. Also matches the matrix code, which expects args to be of the same size
Mon, 31 Mar 2014 10:28:42 +0100 Chris Cannam Merge
Mon, 31 Mar 2014 10:16:02 +0100 Chris Cannam Start on captions
Fri, 28 Mar 2014 13:16:12 +0000 Chris Cannam Work on minValue, maxValue, total
Tue, 25 Mar 2014 10:32:01 +0000 Chris Cannam tiledTo
Mon, 24 Mar 2014 12:14:38 +0000 Chris Cannam entryWiseProduct, another one that should have taken a list
Mon, 24 Mar 2014 11:50:05 +0000 Chris Cannam Make matrix sum also take a list (and add vector random)
Mon, 24 Mar 2014 11:23:28 +0000 Chris Cannam Add repeatedHorizontal/Vertical
Mon, 24 Mar 2014 09:59:27 +0000 Chris Cannam Make vector multiply take a list of vectors, not just two (in symmetry with add); slightly simplify implementations so as to return the arg unmodified if there is only one
Fri, 21 Mar 2014 17:32:48 +0000 Chris Cannam Entry-wise divide
Fri, 21 Mar 2014 16:38:03 +0000 Chris Cannam Abandon ivector, it's less clean and it wasn't any faster