Fix forest_fire notebook for Mesa 4 compatibility

[Vanya-kapoor]

Fixes forest_fire notebook to work with Mesa 4, replacing removed
and deprecated APIs.

Changes:

• Replace deprecated imports (BatchRunner, mesa.datacollection)
  with import mesa and import pandas as pd
• Replace deprecated fire.run_model() with
  while fire.running: fire.step()
• Fix fire.dc → fire.datacollector
• Replace removed BatchRunner with mesa.batch_run()
• Fix DataFrame column access: df.BurnedOut → df["Burned Out"]

Fixes #421
Related: #3132, #3270

[Harshini2411]

Hey, good work on this! The migration approach is solid and you've got the right idea. Just a couple of things to sort out before it's ready to merge:

The big ones:

• DataCollector is never imported. The model still uses it directly but the import cell only has import mesa and import pandas as pd. You'll get a NameError when you run it. Either add from mesa.datacollection import DataCollector or switch to mesa.DataCollector(...) in the model.
• df["BurnedOut"] needs to be df["Burned Out"] (with a space). The reporter key in DataCollector is Burned Out, so that's what batch_run will name the column. Without the space it'll throw a KeyError.

Smaller things:

• There's a # this is fix by vanya comment left in deffuant_weisbuch/model.py worth removing before merging!
• The notebook still has the old error tracebacks in some cells. Could you do a clean run in a Mesa 4 environment and recommit with fresh outputs?

Once those are fixed this should be good to go!

[Vanya-kapoor]

Thanks for your feedback.I have made the changes please see to it

[codebreaker32]

Keep your git history clean

[Vanya-kapoor]

Keep your git history clean

done @codebreaker32

[Vanya-kapoor]

@EwoutH @jackiekazil @catherinedevlin this pr is ready to be merged!