Closed AdamRJensen closed 2 years ago
In the comparison table, I wonder if the error statistics should be calculated after rounding to two decimal places. For example the first row looks a bit wonky the way it is now -- identical (rounded) ASL values, but nonzero relative deviation:
Good catch. I must have done that previously and then mistakenly reverted it. It has been implemented again now.
This PR adds a notebook/documentation section on the validation of the shading calculation method.
The validation is carried out by comparing the annual shading fraction for nine different field layouts presented in Cumpston and Pye (2014).