Changelog

v2.3.1 (2022-11-08)

Fix

  • test_interpolate: Ensure interpolation nodes generally not becoming too small around zero (b2bec0d)

Documentation

  • module diagram: Update module diagram to reflect signals.py module character to resolve #266 (f49a679)

  • module diagram: Update module diagram with signals.py to resolve #266 (aa80faa)

See all commits in this version

v2.3.0 (2022-08-18)

Feature

  • LSIIR: Return the RMS value backwards compatible (see #288) (c5c484e)

Fix

  • test_interpolate: Ensure interpolation nodes not becoming too small in orders of magnitude (f3bf886)

  • test_interpolate: Avoid only zeros as interpolation nodes in strategy for test case generation (2596f24)

  • _compute_stabilized_filter_through_time_delay_iteration: Correct implementation (6286c75)

See all commits in this version

v2.2.0 (2022-04-22)

Feature

  • convolve_unc: Allow 1D array of stdunc as input (ae5335a)

See all commits in this version

v2.1.3 (2022-04-19)

Fix

  • test_ARMA: Increase closeness tolerance (e35e536)

See all commits in this version

v2.1.2 (2022-02-07)

Fix

  • tools: Switch to eigs import from scipy.sparse.linalg for scipy>=1.8.0 (6618278)

See all commits in this version

v2.1.1 (2021-12-18)

Fix

  • LSIIR: Proper init of final_tau (29f2eef)

Documentation

  • Signal: Introduce Signal class into docs (0da9b9d)

  • Python 3.10: Introduce Python 3.10 to the docs (a20384a)

See all commits in this version

v2.1.0 (2021-12-03)

Feature

  • tools: Provide convenience functions to prepare input vectors for DFT and filtering (6d15922)

Documentation

  • examples: Add reference to hydrophone paper (3c7880a)

  • examples: Add regularization example inside DFT best practice (75f6dcc)

See all commits in this version

v2.0.0 (2021-11-05)

Feature

  • Weighted least-squares IIR or FIR filter fit to freq. resp. or reciprocal with uncertainties (8aca955)

  • DWT: Add wavelet transform with online-support (aed3deb)

  • propagate_DWT: Add prototype of wave_rec_realtime (76ca8df)

  • misc: Add buffer-class for realtime applications (d105de2)

  • propagate_DWT: Return the internal state (31fdb19)

  • IIRuncFilter: Always return internal state (175357a)

Fix

  • propagate_filter: Avoid floating point issues with small negative uncertainties via clipping (bbe9d13)

  • FIRuncFilter: Actually perform shifting for fast computation cases (14345c6)

  • FIRuncFilter: Output shifting returns expected covariance matrix (3c6ca41)

  • propagate_DWT: Adjust renamed function (7978c26)

  • imports: Make DWT-methods available from top-level (85165a6)

  • examples: Remove unsed imports (f32d975)

  • examples: Remove unused buffer from speed-comparison-filter (d02a9f3)

  • IIRuncFilter: Take sqrt(Ux[0]) in case of kind=corr (38bdb99)

  • IIRuncFilter: Warn user if Ux is float but kind not diag (47e01f5)

  • IIRuncFilter: Use None as default for Uab (0e7fd18)

  • propagate_filter: Refine error messages (038ef72)

  • example: Remove validate_FIRuncFilter (76d09a2)

  • example: Adjust validate_FIRuncFilter (7469c91)

  • examples: Review validate_DWT_monte_carlo- sort imports- add docstring- fix renamed functions- fix changed signatures\n- apply black (0199dfe)

  • example: Enhance realtime_dwt (14f54fd)

  • model_estimation: Introduce new package model_estimation in preparation of deprecations (627575c)

  • IIRuncFilter: Match default kind with FIRuncFilter (0a0fdfe)

  • propagate_filter: Fix correlated uncertainty formula (70e9375)

  • FIRuncFilter: Set internal state of lfilter (1f60e76)

  • validate_DWT_monte_carlo: Adjust return values of dwt/idwt (4dd601b)

  • test_decomposition_realtime: Adjust concat statement (947ed21)

  • wave_dec_realtime: Missing argument in np.empty (583a7b5)

  • idwt: Remove leftover from debugging (7cca19d)

  • idwt: Adjust boundary conditions (b7788ff)

  • test_dwt: Remove too many unpack values (4b52d67)

Breaking

  • Combine deconvolution.fit_filter and identification.fit_filter into model_estimation.fit_filter and provide access to all functionality via according parameter sets for model_estimation.fit_filter.LSFIR and model_estimation.fit_filter.LSIIR. (8aca955)

  • Rename input parameters t and t_new to x and x_new in PyDynamic.uncertainty.interpolate (918f5bb)

  • Rename fit_sos() to fit_som() because it actually handles second-order models and not second-order-systems. (bc42fd1)

Documentation

  • README: Restyle README and generally improve structure of docs (1409856)

  • Fix some formatting issues resulting in strange looking or misleading info on ReadTheDocs (ab30b4b)

  • Design of a digital deconvolution filter (FIR type): Introduce one more example notebook (c51b98b)

  • uncertainties: Integrate DWT-module to docs (fb7a99a)

  • propagate_DWT: Enhance/prettify docstrings (1fcfc43)

  • IIRuncFilter: Minor adjustments to docstring (475a754)

  • propagate_DWT: Extend module description (a007797)

  • README: Document in README optional dependency installation for Jupyter Notebooks (a59f98d)

  • propagate_filter: Fix IIRuncFilter docstring (e2bd085)

  • propagate_filter: Mention FIR and IIR difference (f6dcd4e)

  • examples: Move validation script to examples (abc0fd9)

  • examples: Include errorbars instead of lines (76d978e)

  • examples: Use latex font and adjust naming (57f4c83)

  • examples: Higher uncertainty, tight layout (58401c3)

  • examples: Refining plot output (3d0e64c)

  • examples: Calculate and highlight 10 biggest coeffs (7e754af)

  • examples: Change order and appearance of plots (3138d51)

  • examples: Realtime_dwt with multi-level-decomposition (6d48ba7)

  • examples: Plot detail coefficients (53ca6f5)

  • examples: Apply dwt to signal (93edef5)

  • examples: Add script to examine realtime Wavelet (eaf13e7)

  • IIRuncFilter: Fix wrong formula reference (0999569)

  • propagate_filter: Adjust return values of IIRuncFilter (02a2350)

  • IIRuncFilter: Describe non-use of b, a, Uab if state (0889475)

  • propagate_filter: Enhance specification of “kind” (ee2062d)

See all commits in this version

v1.11.1 (2021-10-20)

Fix

  • IIRuncFilter: Introduce a missing import of scipy.signal.tf2ss (17fe115)

v1.11.0 (2021-10-15)

Feature

  • plot_vectors_and_covariances_comparison: Introduce function to conveniently compare vectors (e2b3b0c)

  • normalize_vector_or_matrix: Make normalize_vector_or_matrix() publicly available (52b1256)

  • is_2d_square_matrix: Make is_2d_square_matrix() publicly available (e303e6b)

Fix

  • version: Reintroduce version variable into PyDynamic/init.py (0349b09)

Documentation

  • CONTRIBUTING: Mention necessity of installing PyDynamic itself for testing (1571585)

v1.10.0 (2021-09-28)

Feature

  • propagate_DFT: Make some helpers to check for shapes of inputs publicly available (dc97b3f)

Fix

  • fit_som: Apply minor changes to fit_som example to make it executable again (0157fc7)

v1.9.2 (2021-09-21)

Fix

  • PyPI package: Include examples in packag and make sure all tests pass with delivered version (f8326d5)

v1.9.1 (2021-09-15)

Fix

  • DFT_deconv: Replace the imaginary part of H by Y’s imaginary part in one of the equations (a4252dd)

Documentation

  • Introduce Python 3.9 to the docs and actually provide requirements*.txt files (19dcef2)

v1.9.0 (2021-05-11)

Feature

  • interp1d_unc: Add cubic bspline interpolation-kind (f0c6d19)

Documentation

  • interp1d_unc: Add example for kind = “cubic” (8c2ce38)

v1.8.0 (2021-04-28)

Feature

  • propagate_convolution: Convolution with full covariance propagation (299165e)

Documentation

  • propagate_convolution: Add module description (ffafa00)

  • DFT_deconv: Improve wording of docstring for return value (e866aa4)

v1.7.0 (2021-02-16)

Feature

  • FIRuncFilter: Add FIR filter with full covariance support (b937a8b)

Documentation

  • trimOrPad: Enhance docstring (4c0da58)

  • _fir_filter: Adjust docstring and ValueError (090b8f8)

v1.6.1 (2020-10-29)

Fix

  • Flip theta in uncertainty formulas of FIRuncFilter (dd04eea)