fixed some bug in computation of acceptance probability in MetropolisHastings, now works with restrained value ranges too. Also changed sample_local in continuous node to direct sampling so there is no rejection sampling anymore

Had to change linear dependencies in Exponential and Beta Nodes to sigmoid functions on weighted sums. Also introduced a more complex example for Continuous Nodes

Fixed several bugs related to evidence. Hopefully included LinearBeta, still needs testing though. Changed sample_global and sample_local once again in ContinuousNode

examples/MCMC_Continuous.py läuft mit exponentialverteilung->linearGauss, Evidenz als Klassen und Kontinuierliche PoE läuft