check on compatibility of parameter bounds with scaling_uncertainties and other features

[AdityaSavara]

As noted in this pull request,
#88

We have the following possibilities:

Uncertainties scaling can now be turned off, by parameter_estimation_settings['scaling_uncertainties_type'] = 'off'
Uncertainties scaling can now be done by a fixed constant by using a string that is a number, like this: parameter_estimation_settings['scaling_uncertainties_type'] = '1e3'

The default is "std".

However, we now have parameter bounds with
InputParameterPriorValues_upperBounds
InputParameterPriorValues_lowerBounds

This works with "off" for the scaling. It probably does not yet work for any other scaling.

##As of May 4th 2020, this only has been checked for scaling_uncertainties_type = 'off'

[AdityaSavara]

it mightbe useful to put InputParameterPriorValues_upperBounds and the lower bound one in the mcmc sampling directly rather than just the likelihood, but we have to keep it in the likelihood as well in case somebody uses gridsearch or direct sampling.

There are two ways of implementing: One could include it in samples as "zero" probability (logP='-inf') without checking prior and likelihood, or one could even exclude it from the sampling. Both ways make sense to me.

[AdityaSavara]

• The bounds should be made compatible with reducedParameterSpace (done October 9th 2021)

[AdityaSavara]

The bounds check is currently designed that if somebody puts a bound of 0 that excludes 0 from being allowed.

• There should probably be an option to specify whether the bound value should be allowed or disallowed. [low priority, won't implement at the moment]