Jason K. Moore

> My original test for the leg2d code was just to make sure that the answer stayed the same through revisions of the code. Good to know that it was...

Sure, feel free to change that.

Calling .readlines() a file is fast, as done for the event finding. I'm not sure where all of the .readline() methods are being called. ``` In [7]: %time df =...

Here's profiling output of essentially just the DFlowData.clean_data() call: ``` 1317864 function calls (1317260 primitive calls) in 62.622 seconds Ordered by: internal time ncalls tottime percall cumtime percall filename:lineno(function) 128105...

See issue #30 for details on the rtfilter.

So our b vector, in Ax=b, is dense. So maybe sparse linearly solving is fruitless. This http://stackoverflow.com/questions/21507275/why-does-spsolve-function-output-a-vector alludes to that.

lsqr and lsmr seem to be the only ways of solving overdetermined systems (b is considered dense in these algorithms it seems) This can be used by SciPy but may...

@sabrinaabram This is probably easy to fix. The test failure may not be reproducible on your computer but is happening on the Travis ci testing machine. https://travis-ci.org/csu-hmc/GaitAnalysisToolKit/builds/41114121

These seem to be some kind of rotations for various body segments used in the visualizations. They are likely some form of body or space fixed rotations. I've made notes...

@campanawanna Shouldn't your analog relabeling relabel the output of `DFlowData.mocap_column_labels`? I suspect this will having bearing on getting the `write_tsv` method and similar methods to work.