Improve DateRange validation

[maximlt]

Noticed that the DateRange parameter was not checking that the value provided is a tuple, which it should be given that DateRange inherits from Tuple.

I've also changed the error message reported when one of the values is not of the correct type:

DateRange parameter 'a' only takes date/datetime values, not type float.

This is slightly different formatting compared to the messages reported by other Parameters, e.g. with a List:

List parameter None must be a list, not an object of type <class 'str'>.

Notice of type float vs. of type <class 'str'>. If the latter is preferred, I could update the error messages of the other Parameters.

[codecov-commenter]

Codecov Report

Merging #627 (75a2130) into master (b38d3b4) will increase coverage by 0.01%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master     #627      +/-   ##
==========================================
+ Coverage   82.81%   82.82%   +0.01%     
==========================================
  Files           5        5              
  Lines        3037     3039       +2     
==========================================
+ Hits         2515     2517       +2     
  Misses        522      522              

Impacted Files	Coverage Δ
numbergen/__init__.py	80.36% <0.00%> (+0.04%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update b38d3b4...75a2130. Read the comment docs.

[maximlt]

@jlstevens ready for review 🙃

[jlstevens]

Looks good to me!

[jbednar]

Looks good to me, thanks! I agree that it would be nice to make messages more consistent, but don't have a terribly strong opinion between the ways that type information is printed. I think on balance I'd prefer a simpler, more-English like version (e.g. type float) over a syntax-y version (e.g. <class 'float'>).

[philippjfr]

Agreed, consistency would be good but for now this is an improvement.