Implement graph generator functions for common causal graphs

[adam2392]

Is your feature request related to a problem? Please describe.
They should be under pywhy_graphs.generators.*, where each function corresponds to some graph that is generated.

Describe the solution you'd like
The following graphs are very standard in causal inference:

1. napkin graph
2. bow graph
3. iv graph
4. front door graph
5. back door graph

These can be implemented using the ADMG class.

An example pseudocode would be:

def napkin_graph():
     # implementation

Note this differs from networkx-API for generators, because there is no concept of a create_using. Since these are causal graphs with mixed-edges, they would always use a MixedEdgeGraph class.