ProgPy v1.8

ProgPy v1.8

• New Feature Discrete States: Inputs, states, outputs, or performance metrics can now be represented by a discrete state object, which will only exist in a set of defined states. See the Discrete State Notebook for examples of use
• New model: Simplified Battery. This is a simplified version of the BatteryElectroChemEOD model first introduced in the PHM Society Conference ProgPy Tutorial. See Included Models for details
• Support for Python3.13 (with the exception of ProgPy's data driven dependencies due to Tensorflow not supporting this Python version)
• Dropped support of end-of-life Python3.7 and Python3.8
• Improved “ProgPy Short Course”: A series of Jupyter Notebooks designed to help users get started with ProgPy and understand how to use it for prognostics. See https://github.com/nasa/progpy/tree/master/examples
• Various bugfixes and efficiency improvements

ProgPy v1.7.1

HotFix v1.7.1

Fixes issue where datasets do not download correctly.

ProgPy v1.7

Release v1.7

• Started "ProgPy Short Course": A series of Jupyter Notebooks designed to help users get started with ProgPy and understand how to use it for prognostics. See https://github.com/nasa/progpy/tree/master/examples
• Updates to improve composite model:
  • Support setting parameters in composed models using [model].[param] format (e.g., composite_model["model1.Param1"] = 12)
  • Support adding functions to composite. Useful for simple translations
• Prediction and Simulation event strategy. For models with multiple events can now specify if you would like prediction or simulation to end when "first" or "any" of the events are met
• Updates to parameter estimation
  • Users can now estimate nested parameters (e.g., parameters['x0']['a']) using a tuple. For example params=(('x0', 'a'), ...)
  • MSE updated to include a penalty if model becomes unstable (i.e., returns NaN) before minimum threshold. This encourages parameter estimation to converge on parameters for which the model is stable
• Tensorflow no longer installed by default (this is important for users who are space constrained). If you're using the data-driven features install ProgPy like so: pip install progpy[datadriven] or pip install -e '.[datadriven]' (if using local copy)
• Support for Python 3.12
• Removed some warnings
• Various Bugfixes and Performance optimizations

Notes for upgrading:

• If you're using the data-driven features install ProgPy like so: pip install progpy[datadriven] or pip install -e '.[datadriven]' (if using local copy)
• Use "events" keyword instead of "threshold_keys" in simulation

ProgPy v1.6

Release v1.6

• Combined previous prog_models and prog_algs packages into a single package, progpy.
• Added new MixtureOfExperts model, which combines multiple models of the same system into a single model, where only the best of the comprised models will be used at each timestep.
• Added ability to set random seed in GaussianNoiseWrapper, allowing for repeatable experiments
• Various bug fixes and performance improvements

Updating from prog_models/algs v1.5

ProgPy v1.6 combined prog_models and prog_algs into a single package, progpy. To upgrade to 1.6, you will need to download the new progpy package (pip install progpy) and update all imports to use progpy.
For example from prog_models import PrognosticsModel becomes from progpy import PrognosticsModel, and from prog_algs import predictors becomes from progpy import predictors.

Interns

Thank you interns/visiting researchers @aqitya and @mstraut for your contributions to this release