- Marcos and Codas have more verbose output object with more fields.
- Update documentation
- Initial implementation of the Best-Worst method.
- Fixed Best-Worst method's optimization stage inconsistency in Windows.
- Implement IDOCRIW for automatic weight selection
- Fix clos().
- Implement CILOS (The Criterion Impact Loss Method) for determining weights objectively.
- Separate MCDM tests into folders
- The type changed as
comparisonMatrixList::Vector{Matrix{Float64}}
insted ofcomparisonMatrixList::Vector
in ahp (#62).
- Remove legacy normalization in Waspas
- Add documentation for Grey Numbers
- Add citations to operations on Grey Numbers
- Binary comparison operators are redefined as in the citation given
- Add tests for center() and radius() of grey numbers
- Export more inner state variables of critic method.
- Implement nullnormalization for disabling normalization
- Fix docstring of todim method (delete the special characters).
- Implementation of TODIM method as a new MCDM tool.
- ARAS, EDAS, and WASPAS results are now holding more inner-calculation results.
- Add LMAW (Logarithm Methodology of Additive Weights) as a new MCDM tool.
- add OCRA (Operational Competitiveness RAting) a new MCDM tool.
- Add LOPCOW (LOgarithmic Percentage Change-driven Objective Weighting) method for automatic calculation of weights.
- Add normalization option to WPM method
- XYZMethod() now defines the default normalization for the possible copeland calls.
- Test code coverage increased.
- Added isfinite(), isnan(), one() for GreyNumber.
- Added simplify() for GreyNumber.
- Update documentation to method which takes optional normalization argument.
- Add optional normalization to SECA method.
- Add optional normalization to Moora ref and ratio methods.
- Remove Base.show methods of McDMResults. The structure is already readable enough.
- Custom normalization feed for other methods
- Users can now manually feed a normalization method for any MCDM tool.
- Normalization parts of the methods separated as a new module.
- Topsis returns ideal vectors in TopsisResult structure
- Implement SECA method.
- Remove DataFrames dependency from tests and DataEnvelop
- DataEnvelop now works with dictionaries
- Remove redundant nds() method that takes a DataFrame argument.
- Remove redundant functions from Utility.jl after the dependency of DataFrames.jl was removed.
- Update README
- Update Jupyter notebook for example usage
- Reformat source code using JuliaFormatter
- Remove dependency of DataFrames.jl. All of the functions now takes the decision matrix in type of Matrix.
- Dependencies for JuMP and GLPK removed. Requires.jl automatically activates Game and DataEnvelop modules when these two packages are loaded manually by the user.
- Documentation for copeland
- Copeland with vector of methods implemented. With this implementation
met = [
PIVMethod(),
PSIMethod(),
ROVMethod(),
SawMethod(),
VikorMethod(),
WaspasMethod(),
WPMMethod(),
]
result = copeland(df, w, fns, met)
type call is possible
- Remove dependency of
makeminmax()
. Now functions take Vector{F} where {F <: Function}
- Report more result in JMcDMResults
- Add new methods to mcdm()
- Update documentation system and integrate with CI
- More verbose Topsis output with distances to positive and negative ideal solutions.
- Add tests for GreyNumber type
- methods now accept Matrix (in addition to DataFrame)
- PIV (Proximity Indexed Value) method implemented
- Implement MEREC
- Grey Number support for more methods
- game() now returns solution vector for both players
- mcdm() function receives TopsisMethod() as default argument
- yml file updates for testing
- Julia version is updated from ^1.4 to ^1.5
- minor fixes and new tests on mcdm methods
- Grey Numbers and their operations
- Grey Number support for Topsis, Vikor, Aras, Cocoso, Codas, Copras, Edas, Mabac
- Grey Number support for Mairca, Marcos, Moosra, Waspas, Wpm, Saw, Rov, Psi
- Grey Number support for Moora, Electre
- Update documents for CRITIC method
- Update mcdm()
- Modular system
- game() accepts matrix
- Implement MOOSRA method
- Fix game solver
- Implement PSI (Preference Selection Index) method
- Small bug fixes
- Remove the critic method from summary()
- print and summary for ROV method
- Add compat entry for new JuMP version 1.0.0
- Update README
- Fix Vikor
- Update dependencies GLPK and JuMP
- Add new tests for Copeland
- Documentation added for Copeland
- ROV (Range of Value) Method implemented.
- SD method implemented for determining weights.
- fix Copeland.
- add Moora Ratio method with new tests.
- entropy() returns a result even though there are NaNs for some criterion.
- rwrapper.R added so the library is callable from within R
- copeland() method added for combining multiple ordering results.
- Default optimizer is now GLPK (Cbc removed)
- Direction of optimization added for nds()
- New tests added
- Bug in Moore fixed.
- Bug in Marcos fixed.
- Bug in Electre fixed.
- New tests added.
- Tests were divided into several files
- Dependencies upgraded
- On/Off switch for tests.
- New tests.
- Base.show(io:IO, MCDMResult) implementations for pretty printing all of the results
- I() implemented, LinearAlgebra package removed from dependencies.
- mean(), geomean(), std(), and cor() are implemented and StatsBase & Statistics packages are removed
- using keyword replaced by import and only the needed functions are loaded at startup.