Better documentation and use of lognormal parameters

[sbenthall]

The Distribution.LogNormal class takes two parameters, mu and sigma.

These are the mean and standard deviation for the underlying normal distribution:

https://github.com/econ-ark/HARK/blob/master/HARK/distribution.py#L89

https://numpy.org/doc/stable/reference/random/generated/numpy.random.lognormal.html

This could be made more clear in the documentation.

There is a separate method for creating the lognormal from the values of its own mean and standard deviation.

https://github.com/econ-ark/HARK/blob/master/HARK/distribution.py#L202

This latter method is used in update_RiskyDstn:

https://github.com/econ-ark/HARK/blob/master/HARK/ConsumptionSaving/ConsPortfolioModel.py#L209-L249

RiskyDstn is used in the solver.

But the true mean/sigma factory method is not used in get_Risky:

https://github.com/econ-ark/HARK/blob/master/HARK/ConsumptionSaving/ConsPortfolioModel.py#L339-L367

which is used for drawing the shocks in forward simulation.

This is a bit confusing.