Convert objects hierarchy from MATLAB to to python

Question

I am converting a MATLAB script into python. The MATLAB script uses some kind of objects hierarchy for different input variables as in:

parameterA.subparameter1= A1;
parameterA.subparameter2= A2;

parameterB.subparameter1= B1;
parameterB.subparameter2= B2;

Where A1,A2,B1,B2 can all be strings and numbers. I wish to convert this to python and I have used

parameterA.subparameter1= A1
parameterA.subparameter2= A2

parameterB.subparameter1= B1
parameterB.subparameter2= B2

Now I am getting the error:

NameError: name 'parameterA' is not defined

I have tried to initialize them by using either parameterA,parameterB=[],[], which is not good because I want objects and not a list or parameterA,parameterB=set(),set() according to this post as well as some other solution.

Is this at all the correct approach? How do I initialize this structure? Is there a better way to do this?

Answer 1

I would strongly suggest that you use dicts instead. These are efficient and versatile built-in objects in python.

parameterA  = {'subparameter1': A1} # definition
parameterA['subparameter2'] = A2 # update/assignment

parameterB = {'subparameter1': B1, 'subparameter2': B2}

As you can see you can access and modify values using square brackets. If you really insist on using attribute lookup, you could use types.SimpleNamespace:

from types import SimpleNamespace

parameterA = SimpleNamespace(subparameter1=A1)
parameterA.subparameter2 = A2

parameterB = SimpleNamespace(subparameter1=B1, subparameter2=B2)

The main reason I'd stick to dicts is the ease with which you can access keys dynamically. Only use a SimpleNamespace or something similar if you never access attributes programmatically, because your objects are more like a tool for grouping variables together. You can still access attributes programmatically, but it's uglier (and less idiomatic) than dict item lookup.