Checking the zero variance limit was easier than expected. Also, calculating the covariance was easier than expected.

However, I'm really not looking forward to programming the bloody thing. :)

Changes to be committed:
modified: debinningCore/debinning.py
modified: ipython_notebook_slides/debinningWithNewStatistics.ipynb
new file: zeroVarianceLimit.py
new file: zeroVarianceLimit_100.pdf
new file: zeroVarianceLimit_1000.pdf
new file: zeroVarianceLimit_200.pdf
new file: zeroVarianceLimit_2000.pdf
new file: zeroVarianceLimit_500.pdf
new file: zeroVarianceLimit_5000.pdf