Time Series Analysis#

Overview#

The time_series module provides classes and functions for analyzing temporal data, particularly useful for adaptive optics telemetry and signal processing:

  • TimeSeries: Container class for temporal data with time stamps, units, and metadata

  • MultiTimeSeries: Handle multiple synchronized time series

  • Indexer: Index and access time series data efficiently

  • PSD Computer: Power spectral density analysis and temporal frequency analysis

These tools support:

  • Data storage with proper time stamps and physical units

  • Resampling and interpolation

  • Statistical analysis (mean, variance, temporal structure)

  • Frequency domain analysis

  • Data merging and synchronization

Fluent API#

The TimeSeries class provides a modern chainable fluent API for data analysis:

Chainable Properties:

  • Ensemble reductions: ensemble_rms, ensemble_mean, ensemble_std, ensemble_median, ensemble_ptp

  • Time reductions: time_mean, time_std, time_rms, time_median, time_ptp

  • Value extraction: .value property (like pandas)

  • Filtering: filter(**kwargs) and with_times(times)

Example:

# Chainable operations
mean_rms = ts.filter(modes=[2,3,4]).ensemble_rms.time_mean.value

# Bidirectional chaining
long_exposure = ts.time_mean.ensemble_rms.value

See the tutorial notebook for detailed examples.

Legacy API:

Old methods (ensemble_rms(), time_average(), etc.) have been renamed with get_ prefix and marked deprecated. Use the new property-based API for new code.

API Reference#

For detailed API documentation of all submodules, see:

On this page

This Page