Welcome to pactus’ documentation!

Standing from Path Classification Tools for Unifying Strategies, pactus is a Python library that allows testing different classification methods on several trajectory datasets.

It comes with some built-in models and datasets according to the state-of-the-art in trajectory classification. However, it is implemented in an extensible way, so the users can build their own models and datasets.

Overview of the resources available in pactus coupled with an usage example

Note

Code from the example shown above:

from pactus import Dataset, featurizers
from pactus.models import RandomForestModel

dataset = Dataset.animals()
train, test = dataset.split(0.9)

ft = featurizers.UniversalFeaturizer()
model = RandomForestModel(featurizer=ft)
model.train(train, cross_validation=5)

evaluation = model.evaluate(test)
evaluation.show()

Getting Started

• Installation and compatibility
• Getting Support

Examples

• Example 0
  • 1. Setup dependencies
  • 2. Loading Data
  • 3. Inspecting a single trajectory
  • 4. Inspecting a subset of the first trajectories
  • 5. Inspecting the distribution of trajectories on each class
  • 6. Inspecting the lenght distribution of the trajectories in the dataset
• Example 1
  • 1. Setup dependencies
  • 2. Definition of parameters
  • 3. Loading Data
  • 4. Loading the model
  • 5. Training and evaluation
  • 6. References
• Example 2
  • 1. Setup dependencies
  • 2. Definition of parameters
  • 3. Loading Data
  • 4. Loading the model
  • 5. Training and evaluation
  • 6. References
• Example 3
  • 1. Setup dependencies
  • 2. Definition of parameters
  • 3. Loading Data
  • 4. Loading the models
  • 5. Training and evaluation
• Example 4
  • 1. Setup dependencies
  • 2. Definition of parameters
  • 3. Loading Data
  • 4. Loading the model
  • 5. Training and evaluation

Extensibility

• Create new datasets
• Create new models
• Create new featurizers

Contributing

• Increasing Visibility and Engagement
  • Bug reporting
  • Requesting new features
  • Extending pactus source code
  • Adding new examples of usage
  • Increasing pactus visibility
• Adding datasets to pactus
  • 1. Add the dataset to the trajectory-datasets repository
  • 2. Use the dataset
• Adding models to pactus
  • 1. Create the model script
  • 2. Fork the pactus GitHub repository and add the model script
  • 3. Make a pull request

Advanced Resources

• Available datasets
• API Reference
  • Datasets
  • Models

Indices and tables

• Index

• Module Index

• Search Page