Release 1.2

Decreased the progress incrememt when generating the Markov chain from 20% to 5% intervals. This was done to improve the memory requirements.
Fixed a small bug in the 'elrm' and 'update' methods that wrongly repeated the last sampled response vector in the next batch.
Added a check that 'iter' and 'burnIn' are integer values in the 'elrm' and 'update' methods.
Added a check that the Markov chain that is used for conducting inference is atleast of size 1000 in the 'update' method.
