Plot ATT/EPH time series to check jitter

[bpurinton]

One note on the reports - it would be useful to extract and plot time series of the ATT/EPH covariance matrix diagonal magnitude (effectively, std) from the L1B xml. Will potentially help us understand why some products have larger jitter error than others. And the mean values for the full image may be a predictor for the expected pointing precision. Can discuss further when Ben is back. I think we did this in early testing for jitter correction.