udacity / aind-planning Goto Github PK

ModuleNotFoundError: No module named 'aimacode'

This line doesn't work.
sys.path.append(os.path.join(os.path.dirname(parent), "aimacode"))
Replace it by
sys.path.append(os.path.dirname(parent))

In the commit on Apr 19 there was a change named:

• Eliminating pgNode_s.literal attribute (the same information is encoded in the symbol attribute):
  Code change:
  fd65ae7#diff-b7f5258b614ee7cc9148b4b75b539c5fL76

But comments in code on master branch still indicate its presence:
https://github.com/udacity/AIND-Planning/blob/master/my_planning_graph.py#L65

Instance variables calculated:
            literal: expr
                    fluent in its literal form including negative operator if applicable

This leads to confusion when implementing h_levelsum method in my_planning_graph.py as we need to perform goal checks at each level against the same PgNode_s instances. https://github.com/udacity/AIND-Planning/blob/master/my_planning_graph.py#L476

BTW I've even got code review rejected related to this issue as suggestion by the reviewer uses .literal in a conditional.

https://github.com/udacity/AIND-Planning/blob/master/my_planning_graph.py#L87
I think there is a minor bug while showing the PgNode_s at line 87 of my_planning_graph.py, the ~ shouldnt be there for the positive case

print("\n*** ~{}".format(self.symbol))

na1 and na2 have empty precondition. How this test case can assert them competing_needs_mutex() == false, then mutexify(na3, and na4), then assert competing_needs_mutex() == true.

Please consider to add an eq operator to Action class. See pull request 4.
That would resolve problem with your test case in planing project and improve readability of the code.