All notable changes to the "swarmlib" pypi package will be documented in this file.
The format is based on Keep a Changelog,
and this project adheres to Semantic Versioning.
2020-12-16 - v0.14.1
- a bug that caused swarmlib to show an error message instead of the help message when it was invoked without any args (#36)
2020-12-04 - v0.14.0
- new functions by upgrading landscapes to v0.0.11
- a bug that caused levy flights to be performed randomly despite
--seed
was enabled.
2020-11-25 - v0.13.0
- the whale optimization algorithm. After each step the solutions are plotted.
- the upload of release artifacts. Now the bundled pypi packages are added as release artifacts again.
2020-11-17 - v0.12.1
- swarmlib's appearance on pypi. Now the gif renders correctly.
2020-11-16 - v0.12.0
- top level flag
--log-level
to set swarmlib's verbosity (#29). It can be one ofdebug
,info
,warning
,error
,critical
. Contributed by Nikos Koutsovasilis (@nkoutsov) and happily accepted 🚀.
2020-11-10 - v0.11.0
- visualization of the alpha, beta and delta wolves for the grey wolf optimizer (#14).
- top level
--seed
flag which allows to set the random bit generator's initial state for reproducible results.
- the visualization. Now the initial as well as the end-position are shown unanimated as well.
2020-11-07 - v0.10.0
- landscapes as a dependency to enable more benchmark functions (#15). Contributed by Alex F (@alxfmpl). Thanks a lot 🚀.
2020-07-21 - v0.9.0
- grey wolf optimizer (#12). Perform the grey wolf optimization algorithm on one of the selected 2D-functions. Contributed by Nimish Verma (@NimishVerma) and greatly appreciated 🚀.
2020-04-13 - v0.8.1
- a bug that caused the ACO algorithm to fail. Due to other third party packages swarmlib now requires
matplotlib<3.2.0
.
2020-04-07 - v0.8.0
- the artificial bee colony algorithm. After each step the intermediate solution is plotted.
2020-02-17 - v0.7.0
- dark mode. It is enabled via the
--dark
flag.
--continuous
and--interval
flags. Both are now top level flags.- the API of the ant colony optimization.
- the
tsp_file
argument to an option. Now--tsp-file
is optional. By default the built-in burma14 problem is used.
2020-01-25 - v0.6.2
- cuckoo search visualization: when a nest is abandoned / newly generated color its transition differently.
- cuckoo search visualization: now the abandon transition is mapped to the correct nest.
2020-01-24 - v0.6.1
- cuckoo search: Ensure each nest is assigned a cuckoo position in the update step
2020-01-24 - v0.6.0
- the visualization of the firefly algorithm and the cuckoo search. Now they both include velocities.
- the firefly algorithm including its API. Now it replays the same problem if
--continuous
is set. - the
--continuous
flag. It requires no parameter anymore. - the
--two-opt
flag. It requires no parameter anymore.
2020-01-22 - v0.5.0
- a feature that performs particle swarm optimization for one of the provided 2D functions. After each step the intermediate solution is plotted.
- the API. Now the classes
*Problem
can be directly imported from theswarmlib
module.
2020-01-19 - v0.4.1
- pypi tags to enhance the package's discoverability.
2020-01-19 - v0.4.0
- a feature that enables the cuckoo search for one of the provided 2D functions. After each step the intermediate solution is plotted.
2020-01-09 - v0.3.2
- a bug in the firefly algorithm that caused the application to crash when the ackley function was selected.
2019-10-30 - v0.3.1
- a bug in the ACO algorithm that chose the next node by its maximal attractiveness. Now the next node is chosen randomly weighted by its attractiveness
2018-12-18 - v0.3.0
- command line option
--continuous
for the firefly algorithm to indicate, whether the algorithm should run continuously or not - logging for the current best and overall best intensity found of the firefly algorithm
2018-12-14 - v0.2.0
- a feature that enables the firefly algorithm for one of the provided 2D functions. After each step the intermediate solution is plotted.
2018-11-29 - v0.1.0
- a feature that performs 2-opt search once on each partial solution after each iteration. It can be disabled via the argument
--two-opt false
.
- an import bug that caused the application to crash instantly
2018-11-23 - v0.0.1
Initial Release
- a feature that enables solving the Traveling Salesman Problem using the Ant Colony Optimization approach and plots the result afterwards