Filter examples to make smaller

[acwooding]

Add functionality to filter the example datasets down to smaller sizes to run more quickly in CI.

• Add a mock for pd.read_feather
• Generalize run_interactive_examples_script to take a list of context handlers for loading data with parameters (script_file, max_points)
• Add faster tests for cord19 using the example filtering (down to 250k points)
• Add custom cord19 tests using example filtering (again down to 250k points)