preparing for obstable (#79)

* no th unc data in GLP

* mvn

* fix normalisation

* _get_log_likelihood_function_new

* fix type

* finish saving predictions and no_th_unc likelihoods as attributes

* correlated likelihoods as attribute in GLP

* log_likelihood_point

* typo

* multimethod

* no longer properties

* underscore

* change order of selector matrix arguments

* sum log_likelihood_no_theory_uncertainty

* improvements

* compute sm in GL init

* improve docstring

* remove comment

* improvements

Author: alekssmolkovic

pyproject.toml

@@ -18,7 +18,6 @@ dependencies = [
     "jax",
     "wilson",
     "rgevolve",
-    "multipledispatch",
     "networkx",
     "h5py",
 ]