Adam Hughes

jsonify

3

If I replace pickle with json, it should make it possible to use stored states after changing program versions. Not sure if htis just works out of hte box or...

Would also be cool to have a plot that shows full spectrum and another plot (are, referenced etc...) subset, with the lines highlighed in the original plot.

I thought it was in past but now I don't know... Maybe do it this way. Whenever reference is accessed as a property, internally the timespectra check's its own indicies....

http://python-in-the-lab.blogspot.com/2013/12/context-aware-unit-conversion-in-pint.html

Noticed that in pyparty, could comment this out and in fact, if it was not commented out, I got the attributed added to both self.**dict** and self.particle.**dict**. After get testing...

Return series only occurs with attribute access

1

df.x may return a series, but df['x'] will always return a dataframe so maybe I can put this into some code where a series is returned. There are actually several...

If I'm using iunit='A' for example and I subtract baseline, it screws up. See the logic in @property baseline for how to do this. Basically just return to normal, substract,...

Noticed that methods like np.cov() will return a series of values given a timespectra (timeseries actually). Seems like the place to fix this would be in _dfgetattr in TimeSpectra. Tried...

How come this messed up with FiberData March 20 w/ abel's data?

Tested on both **panel** and*\* specstack*\* (save behavior): If I make a timespectra with index from numpy array or Index(), works fine: ``` spec=SpecIndex(np.arange(400, 700,10), unit='nm' ) testdates=date_range(start='3/3/12',periods=30,freq='h') ts=TimeSpectra(abs(np.random.randn(30,30)), columns=testdates,...