Casadi

Latest version: v3.6.7

Safety actively analyzes 714919 Python packages for vulnerabilities to keep your Python projects secure.

Scan your dependencies

Page 14 of 16

1.4.1

Changes

Major

- removed `Mpc`'s `dynamics `property setter; turned it into method `set_dynamics`
- implemented sensitivity for a subset of parameters
- implemented the possibility of instantiating a qrqp solver instead of ipopt
- fixed compatibility for `casadi==3.5.6pre2`

Minor

- refactored `subsevalf` for numpy arrays of symbolic expressions
- improved pickle- and copy-ability of classes with slots
- removed redundant fields in `objective.py`
- made `array2cs` faster
- better type hinting, docs, testing and examples

1.4.0

Changes

Major

- removed `dict2struct` and its use: now most of the properties return a reference to internal dictionaries of variables, parameters, constraints, etc.
- refactored internal files: the same functions can still be access, but their source files have been rearranged
- `MultistartNlp.solve_multi` and `wrappers.NlpScaling.solve_multi` now accept single dictionaries of parameters and variables for multiple scenarions
- `wrappers.NlpScaling` now stores both unscaled and scaled variables and parameters to allow easier (un)scaling of expressions

Minor

- remove `np_random` and `jaggedstack` from utilities (were unused)
- changed initial state names in `wrappers.Mpc` from `{statename}` to `{statename}_0`
- changed argument name `dim` to `size` in some methods in `wrappers.Mpc`
- `csnlp.core.scaling.Scaler` now inherits `dict` instead of `UserDict` for better type hints
- better docstrings, type hints, examples and tests

1.3.2

Changes

- implemented in `wrappers.Mpc` properties `ns`, `na`, `nd`, `nslacks`

1.3.1

Changes

- `Nlp` and `MultistartNlp` support now `__call__`
- refactored `Nlp` for improved maintainability
- improved `Nlp.debug` and utility `invalidate_cache`
- better type hints
- fixed broken link to examples folder in README

1.3.0

Changes

- implemented wrapper `NlpScaling` for nlp scaling (for faster and better convergence)
- made nlp and wrappers **pickleable** and **deepcopy**-able
- added `Nlp.is_wrapped` for checking whether the nlp instance is wrapped in a given type of wrapper
- better tests and code coverage

1.2.3

Changes

- improved CI
- loosened dependency on CasADi

Page 14 of 16

© 2025 Safety CLI Cybersecurity Inc. All Rights Reserved.