trajectory_to_dataframe

viz.trajectory_to_dataframe(trajectory, normalize_position=True)

Convert trajectory components to a tidy pandas DataFrame.

This helper creates a long-format DataFrame suitable for plotting with seaborn, plotly, or custom matplotlib code. Each row represents one position along one trajectory component.

Parameters

Name	Type	Description	Default
trajectory	TrajectoryComponents	The trajectory data to convert.	required
normalize_position	bool	If True (default), position values are normalized to [0, 1] representing relative narrative time. If False, use integer indices (0, 1, 2, …).	True

Returns

Name	Type	Description
	pandas.DataFrame	Tidy DataFrame with columns: - position : float or int, location along narrative - component : str, one of “raw”, “rolling”, or “dct” - value : float, sentiment score at this position

Examples

>>> from moodswing import prepare_trajectory, trajectory_to_dataframe
>>> import matplotlib.pyplot as plt
>>> import seaborn as sns
>>>
>>> # Create trajectory
>>> trajectory = prepare_trajectory(
...     scores,
...     rolling_window=50,
...     dct_transform=DCTTransform(low_pass_size=10)
... )
>>>
>>> # Convert to DataFrame
>>> df = trajectory_to_dataframe(trajectory)
>>>
>>> # Plot with seaborn
>>> sns.lineplot(data=df, x='position', y='value', hue='component')
>>> plt.show()
>>>
>>> # Filter to specific components
>>> df_smooth = df[df['component'].isin(['rolling', 'dct'])]
>>> df_smooth.pivot(index='position', columns='component', values='value').plot()  #
>>>
>>> # Export for further analysis
>>> df.to_csv('sentiment_trajectory.csv', index=False)