Why is the default of many keyword arguments set to None? And what happens then?

[mbakker7]

In many flopy classes, many keyword arguments are set to None. If they are not set by the user, I presume there is some behind-the-scene code that sets these parameters. I find it unclear what is set or how to find out what is set, so it is difficult to decide whether a keyword argument should be specified.

Simple example (but there are thousands):
In mf6.modflow.ModflowGwfwel one of the input arguments is print_input=None. Which I presume is a boolean. Now if I don't specify this parameter, what happens? Is it set to True or False?

[wpbonelli]

Now if I don't specify this parameter, what happens? Is it set to True or False?

In this case, it's set to None. And on write_simulation() the PRINT_INPUT option is omitted from the input file.

I hope (though am not sure — also finding it daunting to determine whether) this logic is universal for keyword (i.e. boolean) variables.

Another gripe. With print_input specified, it ends up an MFScalar, not a boolean, so e.g. if you pass print_input=True you get

> wel.print_input
{internal}
(True)

but

> wel.print_input == True
False

We have this sort of thing prominently in mind as we prototype v4.

[mbakker7]

Thanks @wpbonelli. And glad to hear v4 may have some of these issues resolved.

If I understand you correctly, if a boolean argument is None it means it is not written to the file. In this case this means None defaults to False. Not sure it always works that way (see my discussion #2454, where I am pretty confused). It would be great if in v4 the default for a boolean is True or False and never None. But that is just my opinion.

[wpbonelli]

It would be great if in v4 the default for a boolean is True or False and never None. But that is just my opinion.

Agreed on this.

Will do some digging on xt3d_rhs. I see we have no tests or example models using it..