FelooPy: An integrated optimization environment (IOE) for automated operations research (AutoOR) in Python.
Project description
Introduction
FelooPy (pronounced /fɛlupaɪ/) is an integrated optimization environment (IOE) designed as a decision optimization hub. It involves the use of automated operations research (AutoOR) methods and techniques to identify feasible solutions that lead to logical decisions with the optimal (best possible) outcomes based on given or learnable policies. This Python library simplifies and enhances the use of existing and originally developed modeling, algorithm development, and analytical tools for decision-making within simulated systems, industries, and supply chains. FelooPy is an acronym alluding to an operations research scientist (a.k.a, decision scientist) in pursuit of feasible solutions, logical decisions, and optimal outcomes by optimization in Python. Additionally, it alludes to the concept of loops in computer programming and the floppy disk, symbolizing computing and memory efficiency. The development of FelooPy, which started in September 2022, continues under the MIT license.
Overview:
Learn more:
Features
FelooPy supports the following mathematical structure-based classification of optimization problems:
Display as a list
- Numerical optimization
- Linear Programming (LP)
- [Unconstrained] Linear Programming (ULP, or LP)
- [Constrained] Linear Programming (CLP, or LP)
- General Linear Programming (GLP, or LP)
- Non-Linear Programming (NLP)
- with non-linear objectives
- [Unconstrained] Quadratic Programming (UQP, or QP)
- [Constrained] Quadratic Programming (CQP, or QP)
- with non-linear constraints
- Second Order Cone Programming (SOCP)
- with non-linear objectives and constraints
- General Non-Linear Programming (GNLP)
- with non-linear objectives
- Linear Programming (LP)
- Combinatorial optimization
- Integer Programming (IP)
- Pure Integer Linear Programming (PILP)
- [Unconstrained] Pure Integer Linear Programming (UPILP, or PILP)
- [Constrained] Pure Integer Linear Programming (CPILP, or PILP)
- Pure Integer Non-Linear Programming (PINLP)
- with non-linear objectives
- [Unconstrained] Integer Quadratic Programming (UIQP, IQP, or QUIO)
- [Unconstrained] Binary Quadratic Programming (UBQP, BQP, or QUBO)
- [Constrained] Integer Quadratic Programming (CIQP, IQP, or QUIO)
- [Constrained] Binary Quadratic Programming (CBQP, BQP, or QUBO)
- with non-linear constraints
- with non-linear objectives and constraints
- General Pure Integer Non-Linear Programming (GPINLP)
- with non-linear objectives
- Pure Integer Linear Programming (PILP)
- Mixed Integer Programming (MIP)
- Mixed Integer Linear Programming (MILP)
- [Unconstrained] Mixed Integer Linear Programming (UMILP, or MILP)
- [Constrained] Mixed Integer Linear Programming (CMILP, or MILP)
- Mixed Integer Non-Linear Programming (MINLP)
- with non-linear objectives
- [Unconstrained] Mixed Integer Quadratic Programming (UMIQP, or MIQP)
- [Constrained] Mixed Integer Quadratic Programming (CMIQP, or MIQP)
- with non-linear constraints
- with non-linear objectives and constraints
- General Mixed Integer Non-Linear Programming (GMINLP, or GMIP)
- with non-linear objectives
- Mixed Integer Linear Programming (MILP)
- Integer Programming (IP)
Credit: Keivan Tafakkori
Display as a graph
graph LR
CLASS["FelooPy"] --> SUBCLASS1["Numerical Optimization"]
CLASS["FelooPy"] --> SUBCLASS2["Combinatorial Optimization"]
SUBCLASS1["Numerical Optimization"] --> A["LP"]
SUBCLASS1["Numerical Optimization"] --> B["NLP"]
SUBCLASS2["Combinatorial Optimization"] --> C["IP"]
SUBCLASS2["Combinatorial Optimization"] --> D["MIP"]
A["LP"] --> A1["ULP, or LP"]
A["LP"] --> A2["CLP, or LP"]
A["LP"] --> A3["GLP, or LP"]
B["NLP"] --> B1["NLO"]
B1["NLO"] --> B11["UQP, or QP"]
B1["NLO"] --> B12["CQP, or QP"]
B["NLP"] --> B2["NLC"]
B2["NLC"] --> B21["SOCP"]
B["NLP"] --> B3["NLOC"]
B3["NLOC"] --> B31["GNLP"]
C["IP"] --> C1["PILP"]
C1["PILP"] --> C11["UPILP, or PILP"]
C1["PILP"] --> C12["CPILP, or PILP"]
C["IP"] --> C2["PINLP"]
C2["PINLP"] --> C21["NLO"]
C21["NLO"] --> C211["UIQP, IQP, or UQIO"]
C21["NLO"] --> C212["UBQP, BQP, or UBIO"]
C21["NLO"] --> C213["CIQP, IQP, or CQIO"]
C21["NLO"] --> C214["CBQP, BQP, or CBIO"]
C2["PINLP"] --> C22["NLC"]
C2["PINLP"] --> C23["NLOC"]
C23["NLOC"] --> C231["GPINLP"]
D["MIP"] --> D1["MILP"]
D1["MILP"] --> D11["UMILP, or MILP"]
D1["MILP"] --> D12["CMILP, or MILP"]
D["MIP"] --> D2["MINLP"]
D2["MINLP"] --> D21["NLO"]
D21["NLO"] --> D211["UMIQP"]
D21["NLO"] --> D212["CMIQP"]
D2["MINLP"] --> D22["NLC"]
D2["MINLP"] --> D23["NLOC"]
D23["NLOC"] --> D231["GMINLP, or GMIP"]
Credit: Keivan Tafakkori
FelooPy supports the following expert-based classification of decision-making problems:
Display as a list
-
Multi-Attribute Decision-Making (MADM)
- Weighting methods
- without a decision-making matrix
- with a decision-making matrix
- Ranking methods
- Compensatory methods
- Scoring methods
- Compromising methods
- Non-compensatory methods
- Conjunctive satisfying methods
- Lexicographic methods
- Outranking methods
- Compensatory methods
- Weighting methods
-
Group Decision-Making (GDM)
Credit: Keivan Tafakkori
Display as a graph
graph LR
CLASS["FelooPy"] --> SUBCLASS1["MADM"]
CLASS["FelooPy"] --> SUBCLASS2["GDM"]
SUBCLASS1["MADM"] --> A["Weighting"]
SUBCLASS1["MADM"] --> B["Ranking"]
A["Weighting"] --> A1["without decision matrix"]
A["Weighting"] --> A2["with decision matrix"]
B["Ranking"] --> B1["Compensatory"]
B["Ranking"] --> B2["Non-compensatory"]
B1["Compensatory"] --> B11["Scoring"]
B1["Compensatory"] --> B12["Compromising"]
B2["Non-compensatory"] --> B21["Conjunctive satisfying"]
B2["Non-compensatory"] --> B22["Lexicographic"]
B2["Non-compensatory"] --> B23["Outranking"]
Credit: Keivan Tafakkori
Installation
For a quick installation with a classic support of interfaces and solvers, you may use the pip package manager (please refer to this link to install, update, or get one) as follows:
pip install -U feloopy[stock]
However, as some users might prefer a dedicated version, the following lists the available variants of FelooPy:
Core variant
This variant installs the base package without any additional features. It installs FelooPy with its common dependencies.
pip install -U feloopy
Free variants
-
picovariant:This variant installs the base package without any additional features. It is the same as the core variant. It installs FelooPy with its common dependencies.
pip install -U feloopy[pico]
-
nanovariant:This variant includes a small set of additional features. It installs FelooPy with its common dependencies and the
pymprogpackage.pip install -U feloopy[nano]
-
microvariant:This variant includes a moderate set of additional features. It installs FelooPy with its common dependencies and the
pymprog,gekko, andmealpypackages.pip install -U feloopy[micro]
-
minivariant:This variant includes a large set of additional features. It installs FelooPy with its common dependencies and the
pymprog,gekko,mealpy,ortools, andcvxpypackages.pip install -U feloopy[mini]
-
fullvariant:This variant includes all available features. It installs FelooPy with its common dependencies and the
pymprog,gekko,mealpy,ortools,cvxpy,pymoo, andpydecisionpackages.pip install -U feloopy[full]
-
stockvariant:This variant includes all interface packages. It installs FelooPy with its common dependencies and the
gekko,ortools,pulp,pyomo,pymprog,picos,linopy,cvxpy,mip,mealpy,pydecision,rsome,pymoo,niapy, andpygadpackages.pip install -U feloopy[stock]
Commercial variants
plus_gurobi variant:
This variant includes the Gurobi solver. It requires a valid Gurobi license. Refer to the Gurobi website for more information.
pip install -U feloopy[plus_gurobi]
plus_cplex variant:
This variant includes the CPLEX solver. It requires a valid CPLEX license. Refer to the CPLEX website for more information.
pip install -U feloopy[plus_cplex]
plus_xpress variant:
This variant includes the Xpress solver. It requires a valid Xpress license. Refer to the Xpress website for more information.
pip install -U feloopy[plus_xpress]
plus_copt variant:
This variant includes the COPT solver. It requires a valid COPT license. Refer to the COPT website for more information.
pip install -U feloopy[plus_copt]
Non-compatible variants
only_cylp variant:
This variant includes the CyLP solver. It requires a valid CyLP installation. Refer to this link for more information.
pip install -U feloopy[plus_cylp]
only_linux variant:
This variant includes additional features for Linux-based distros. It installs FelooPy with its common dependencies and the pymultiobjective package.
pip install -U feloopy[only_linux]
Complete variants
-
hypervariant:This variant includes all interface and solver packages. It installs FelooPy with its common dependencies and the
gekko,ortools,pulp,pyomo,pymprog,picos,linopy,cvxpy,mip,mealpy,pydecision,rsome,pymoo,niapy,pygad,cplex,docplex,xpress,gurobipy,cylp, andcoptpypackages.pip install -U feloopy[hyper]
-
megavariant:This variant includes all interface and solver packages. It installs a complete FelooPy with all its dependencies, regardless of whether they are compatible with your operating system, or not.
pip install -U feloopy[mega]
Dev variants
To contribute to the project, support the developer with pull requests, and to get the latest updates, you can install a development variant as follows:
pip install -U git+https://github.com/ktafakkori/feloopy.git#egg=feloopy[variant_name]
or
git clone https://github.com/ktafakkori/feloopy.git
cd feloopy
pip install .[variant_name]
or
!git clone https://github.com/ktafakkori/feloopy.git
%cd feloopy
!pip install .[variant_name]
where variant_name is one of the above variants. (please refer to this link to install, update, or get git.)
Usage
-
Command line interface
To verify FelooPy's command line interface (CLI) accessibility, open a bash, activate the virtual environment or use the global environment with FelooPy installed, and execute either of the following commands:
feloopy -vor
flp -vDetails
Next, you can create your optimization project:
flp project --name=test
This command opens a GUI interface to assist with placing the project folder and prints the project directory for you to navigate using the
cdcommand.The FelooPy's optimization project structure is as follows:
test ├── data │ ├── final │ ├── processed │ └── raw ├── debug.ipynb ├── log.txt ├── main.py ├── modules │ └── __init__.py └── results ├── figures ├── tables └── texts
Note that at specific project progress levels, you can create backups from the project root using:
flp backupThis generates a backup file, preserving the project progress up to a specific date and time, as illustrated below:
test ├── backups │ └── bkp-on-2023-12-05-at-21-00-00.zip ├── data │ ├── final │ ├── processed │ └── raw ├── debug.ipynb ├── log.txt ├── main.py ├── modules │ └── __init__.py └── results ├── figures ├── tables └── texts
To recover to a specific project state, use the following command from the project root:
flp recoverTo clean residuals, execute the following command from the project root:
flp clean -
Quick start
Exact optimization
- Note : Implementing this example at least requires installing the
feloopy[nano]variant.
from feloopy import * # Define a model m = model('exact', 'target_model_name', 'pymprog') # Define variables x = m.bvar(name='x', dim=0) y = m.pvar(name='y', dim=0, bound = [0, 10]) # Define constraints m.con(x + y <= 1, label='c1') m.con(x - y >= 1, label='c2') # Define an objective m.obj(x + y) # Solve the model m.sol(['max'], 'glpk') # Report the results m.report()
Display the output
╭─ FelooPy v0.2.8 ───────────────────────────────────────────────────────────────╮ │ │ │ Date: 2023-12-04 Time: 00:00:00 │ │ Interface: pymprog Solver: glpk │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Model ────────────────────────────────────────────────────────────────────────╮ │ │ │ Name: target_model_name │ │ Feature: Class: Total: │ │ Positive variable 1 1 │ │ Binary variable 1 1 │ │ Total variables 2 2 │ │ Objective - 1 │ │ Constraint 2 2 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Solve ────────────────────────────────────────────────────────────────────────╮ │ │ │ Method: exact Objective value │ │ Status: max │ │ optimal 1.00 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Metric ───────────────────────────────────────────────────────────────────────╮ │ │ │ CPT (microseconds) 347.00 │ │ CPT (hour:min:sec) 00:00:00 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Decision ─────────────────────────────────────────────────────────────────────╮ │ │ │ x = 1.0 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯
Heuristic optimization
from feloopy import * def instance(X): # Define model instance m = model('heuristic', 'representor_model_name', 'feloopy', X) # Define variables for the model instance x = m.bvar(name='x', dim=0) y = m.pvar(name='y', dim=0, bound = [0, 10]) # Define constraints for the model instance m.con(x + y |l| 1, label='c1') m.con(x - y |g| 1, label='c2') # Define an objective for the model instance m.obj((x-1)**2+(y-1)**2) # Solve the model instance m.sol(['max'], 'ga', {'epoch': 1000, 'pop_size': 100}) return m[X] # Make the main model m = make_model(instance) # Solve the model m.sol(penalty_coefficient=10) # Report the results m.report()
Display the output
╭─ FelooPy v0.2.8 ───────────────────────────────────────────────────────────────╮ │ │ │ Date: 2023-12-04 Time: 00:00:00 │ │ Interface: feloopy Solver: ga │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Model ────────────────────────────────────────────────────────────────────────╮ │ │ │ Name: representor_model_name │ │ Feature: Class: Total: │ │ Positive variable 1 1 │ │ Binary variable 1 1 │ │ Total variables 2 2 │ │ Objective - 1 │ │ Constraint 2 2 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Solve ────────────────────────────────────────────────────────────────────────╮ │ │ │ Method: heuristic Objective value │ │ Status: max │ │ feasible (constrained) 1.00 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Metric ───────────────────────────────────────────────────────────────────────╮ │ │ │ CPT (microseconds) 1.31e+06 │ │ CPT (hour:min:sec) 00:00:01 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Decision ─────────────────────────────────────────────────────────────────────╮ │ │ │ x = [1.] │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯
Convex optimization
- Note : Implementing this example at least requires installing the
feloopy[mini]variant.
from feloopy import * # Define a model m = model('convex', 'convex_model_name', 'cvxpy') # Define variables x = m.ftvar(name='x',shape = 0) # Define constraints m.con(x <= 1, label='c1') m.con(x >= 1, label='c2') # Define an objective m.obj((x-4)**2) # Solve the model m.sol(['min'], 'ecos') # Report the results m.report()
Display the output
╭─ FelooPy v0.2.8 ───────────────────────────────────────────────────────────────╮ │ │ │ Date: 2023-12-04 Time: 00:00:00 │ │ Interface: cvxpy Solver: ecos │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Model ────────────────────────────────────────────────────────────────────────╮ │ │ │ Name: convex_model_name │ │ Feature: Class: Total: │ │ Free variable 1 1 │ │ Total variables 1 1 │ │ Objective - 1 │ │ Constraint 2 2 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Solve ────────────────────────────────────────────────────────────────────────╮ │ │ │ Method: convex Objective value │ │ Status: min │ │ optimal 9.00 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Metric ───────────────────────────────────────────────────────────────────────╮ │ │ │ CPT (microseconds) 1.06e+04 │ │ CPT (hour:min:sec) 00:00:00 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Decision ─────────────────────────────────────────────────────────────────────╮ │ │ │ x = 1.000000005186514 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯
Constraint optimization
- Note : Implementing this example at least requires installing the
feloopy[mini]variant.
from feloopy import * # Define a model m = model('constraint', 'satisfaction_model_name', 'ortools_cp') # Define variables x = m.bvar(name='x', dim=0) y = m.ivar(name='y', dim=0, bound = [0, 10]) # Define constraints m.con(x + y <= 1, label='c1') m.con(x - y >= 1, label='c2') # Define an objective m.obj(x + y) # Solve the model m.sol(['max'], 'ortools') # Report the results m.report()
Display the output
╭─ FelooPy v0.2.8 ───────────────────────────────────────────────────────────────╮ │ │ │ Date: 2023-12-04 Time: 00:00:00 │ │ Interface: ortools_cp Solver: ortools │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Model ────────────────────────────────────────────────────────────────────────╮ │ │ │ Name: satisfier_model_name │ │ Feature: Class: Total: │ │ Binary variable 1 1 │ │ Integer variable 1 1 │ │ Total variables 2 2 │ │ Objective - 1 │ │ Constraint 2 2 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Solve ────────────────────────────────────────────────────────────────────────╮ │ │ │ Method: constraint Objective value │ │ Status: max │ │ optimal 1.00 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Metric ───────────────────────────────────────────────────────────────────────╮ │ │ │ CPT (microseconds) 3.65e+04 │ │ CPT (hour:min:sec) 00:00:00 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Decision ─────────────────────────────────────────────────────────────────────╮ │ │ │ x = 1 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯
Multi-objective optimization
- Note : Implementing this example at least requires installing the
feloopy[full]variant.
from feloopy import * def instance(X): # Define model instance m = model('heuristic','representor_model_name','pymoo', X) # Define variables for the model instance x = m.pvar(name = 'x', dim = [2], bound = [-1000,1000]) # Define objectives for the model instance m.obj(x[...,0]**2 + x[...,1]**2) m.obj((x[...,0]-2)**2 + (x[...,1]-2)**2) # Solve the model instance m.sol(['min','min'], 'ns-ga-ii', {'n_gen': 100}, obj_id='all') return m[X] # Make the main model m = make_model(instance) # Solve the model m.sol() # Report the results m.report()
Display the output
╭─ FelooPy v0.2.8 ───────────────────────────────────────────────────────────────╮ │ │ │ Date: 2023-12-04 Time: 00:00:00 │ │ Interface: pymoo Solver: ns-ga-ii │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Model ────────────────────────────────────────────────────────────────────────╮ │ │ │ Name: representor_model_name │ │ Feature: Class: Total: │ │ Positive variable 1 2 │ │ Total variables 1 2 │ │ Objective - 2 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Solve ────────────────────────────────────────────────────────────────────────╮ │ │ │ Method: heuristic Objective value │ │ Status: min min │ │ feasible (unconstrained) 0.00 7.83 │ │ feasible (unconstrained) 7.94 0.00 │ │ feasible (unconstrained) 0.00 7.70 │ │ feasible (unconstrained) 0.08 6.82 │ │ feasible (unconstrained) 2.50 1.57 │ │ feasible (unconstrained) 3.53 0.91 │ │ feasible (unconstrained) 0.70 4.01 │ │ feasible (unconstrained) 0.30 5.26 │ │ feasible (unconstrained) 2.38 1.72 │ │ feasible (unconstrained) 5.17 0.31 │ │ feasible (unconstrained) 2.02 1.99 │ │ feasible (unconstrained) 6.78 0.05 │ │ feasible (unconstrained) 0.63 4.18 │ │ feasible (unconstrained) 5.63 0.22 │ │ feasible (unconstrained) 3.91 0.72 │ │ feasible (unconstrained) 1.28 2.96 │ │ feasible (unconstrained) 1.06 3.46 │ │ feasible (unconstrained) 5.42 0.25 │ │ feasible (unconstrained) 2.62 1.50 │ │ feasible (unconstrained) 0.58 4.28 │ │ feasible (unconstrained) 6.28 0.11 │ │ feasible (unconstrained) 6.14 0.15 │ │ feasible (unconstrained) 0.26 5.39 │ │ feasible (unconstrained) 0.36 5.07 │ │ feasible (unconstrained) 3.42 0.98 │ │ feasible (unconstrained) 2.96 1.23 │ │ feasible (unconstrained) 1.18 3.19 │ │ feasible (unconstrained) 3.69 0.83 │ │ feasible (unconstrained) 1.54 2.53 │ │ feasible (unconstrained) 4.49 0.50 │ │ feasible (unconstrained) 1.77 2.25 │ │ feasible (unconstrained) 2.15 1.88 │ │ feasible (unconstrained) 0.18 5.85 │ │ feasible (unconstrained) 0.20 5.70 │ │ feasible (unconstrained) 3.76 0.79 │ │ feasible (unconstrained) 1.47 2.61 │ │ feasible (unconstrained) 4.63 0.46 │ │ feasible (unconstrained) 6.45 0.10 │ │ feasible (unconstrained) 6.97 0.04 │ │ feasible (unconstrained) 7.35 0.01 │ │ feasible (unconstrained) 3.30 1.03 │ │ feasible (unconstrained) 3.06 1.18 │ │ feasible (unconstrained) 0.42 4.80 │ │ feasible (unconstrained) 5.30 0.28 │ │ feasible (unconstrained) 5.78 0.21 │ │ feasible (unconstrained) 2.30 1.78 │ │ feasible (unconstrained) 0.97 3.72 │ │ feasible (unconstrained) 0.85 3.75 │ │ feasible (unconstrained) 4.82 0.42 │ │ feasible (unconstrained) 1.83 2.18 │ │ feasible (unconstrained) 7.20 0.02 │ │ feasible (unconstrained) 1.14 3.32 │ │ feasible (unconstrained) 5.98 0.17 │ │ feasible (unconstrained) 7.73 0.00 │ │ feasible (unconstrained) 2.85 1.32 │ │ feasible (unconstrained) 1.62 2.42 │ │ feasible (unconstrained) 0.77 3.91 │ │ feasible (unconstrained) 4.24 0.59 │ │ feasible (unconstrained) 4.99 0.38 │ │ feasible (unconstrained) 0.49 4.54 │ │ feasible (unconstrained) 5.90 0.19 │ │ feasible (unconstrained) 0.11 6.39 │ │ feasible (unconstrained) 0.09 6.75 │ │ feasible (unconstrained) 0.09 6.58 │ │ feasible (unconstrained) 0.52 4.44 │ │ feasible (unconstrained) 7.08 0.03 │ │ feasible (unconstrained) 1.42 2.73 │ │ feasible (unconstrained) 1.67 2.36 │ │ feasible (unconstrained) 0.16 5.98 │ │ feasible (unconstrained) 0.12 6.19 │ │ feasible (unconstrained) 3.15 1.11 │ │ feasible (unconstrained) 2.20 1.84 │ │ feasible (unconstrained) 1.94 2.07 │ │ feasible (unconstrained) 7.51 0.01 │ │ feasible (unconstrained) 0.46 4.65 │ │ feasible (unconstrained) 1.35 2.84 │ │ feasible (unconstrained) 4.75 0.44 │ │ feasible (unconstrained) 6.60 0.07 │ │ feasible (unconstrained) 7.57 0.01 │ │ feasible (unconstrained) 0.77 3.82 │ │ feasible (unconstrained) 3.20 1.08 │ │ feasible (unconstrained) 1.91 2.11 │ │ feasible (unconstrained) 0.37 4.96 │ │ feasible (unconstrained) 4.98 0.38 │ │ feasible (unconstrained) 2.71 1.40 │ │ feasible (unconstrained) 0.11 6.28 │ │ feasible (unconstrained) 0.99 3.56 │ │ feasible (unconstrained) 7.84 0.00 │ │ feasible (unconstrained) 0.10 6.51 │ │ feasible (unconstrained) 0.23 5.54 │ │ feasible (unconstrained) 2.80 1.35 │ │ feasible (unconstrained) 1.29 2.91 │ │ feasible (unconstrained) 0.99 3.60 │ │ feasible (unconstrained) 4.04 0.67 │ │ feasible (unconstrained) 0.43 4.72 │ │ feasible (unconstrained) 6.59 0.09 │ │ feasible (unconstrained) 4.36 0.55 │ │ feasible (unconstrained) 1.40 2.78 │ │ feasible (unconstrained) 4.37 0.54 │ │ feasible (unconstrained) 4.12 0.64 │ │ payoff 0 0.00 7.83 │ │ payoff 1 7.94 0.00 │ │ max 7.94 7.83 │ │ ave 2.90 2.40 │ │ std 2.42 2.20 │ │ min 0.00 0.00 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Metric ───────────────────────────────────────────────────────────────────────╮ │ │ │ CPT (microseconds) 1.58e+06 │ │ CPT (hour:min:sec) 00:00:01 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Decision ─────────────────────────────────────────────────────────────────────╮ │ │ │ x[0] = [0.028609389086682313, 1.9859341076320334] │ │ x[1] = [0.006622432714834758, 1.9993559166051682] │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯
Multi-attribute decision-making
- Note : Implementing this example at least requires installing the
feloopy[mini]variant.
from feloopy import * # Define a model m = madm('ahp','ahp_model', 'pydecision') # Define criteria pairwise comparison matrix m.add_cpcm([ [1 , 1/3, 1/5, 1 , 1/4, 1/2, 3 ], [3 , 1 , 1/2, 2 , 1/3, 3 , 3 ], [5 , 2 , 1 , 4 , 5 , 6 , 5 ], [1 , 1/2, 1/4, 1 , 1/4, 1 , 2 ], [4 , 3 , 1/5, 4 , 1 , 3 , 2 ], [2 , 1/3, 1/6, 1 , 1/3, 1 , 1/3], [1/3, 1/3, 1/5, 1/2, 1/2, 3 , 1 ] ]) # Define solve method m.sol() # Report the results m.report(show_tensors=True)
Display the output
╭─ FelooPy v0.2.8 ───────────────────────────────────────────────────────────────╮ │ │ │ Date: 2023-12-04 Time: 00:00:00 │ │ Interface: pydecision Solver: ahp_method │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Model ────────────────────────────────────────────────────────────────────────╮ │ │ │ Name: ahp_model │ │ cpm_defined │ │ Number of criteria: 7 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Solve ────────────────────────────────────────────────────────────────────────╮ │ │ │ Method: ahp_method │ │ Status: feasible (solved) │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Metric ───────────────────────────────────────────────────────────────────────╮ │ │ │ Inconsistency: 0.1094 │ │ CPT (microseconds): 228 │ │ CPT (hour:min:sec): 00:00:00 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Decision ─────────────────────────────────────────────────────────────────────╮ │ │ │ wv = [0.072 , 0.1518, 0.3668, 0.0734, 0.2064, 0.0612, 0.0685] │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯
Uncertain optimization
- Note : Implementing this example at least requires installing the
feloopy[full]variant.
from feloopy import * # Define a model m = model('uncertain', 'mean_varience_portfolio', 'rsome_ro') # Define parameters n = 150 i = np.arange(1, n+1) p = 1.15 + i*0.05/150 sigma = 0.05/450 * (2*i*n*(n+1))**0.5 phi = 5 Q = np.diag(sigma**2) # Define variables x = m.ptvar('x', [n]) # Define an objective m.obj(p@x - phi*m.quad(x, Q)) # Define constraints m.con(x.sum() == 1) # Solve the model m.sol(['max'], 'ecos') # Report the results m.report()
Display the output
╭─ FelooPy v0.2.8 ───────────────────────────────────────────────────────────────╮ │ │ │ Date: 2023-12-04 Time: 00:00:00 │ │ Interface: rsome_ro Solver: ecos │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Model ────────────────────────────────────────────────────────────────────────╮ │ │ │ Name: mean_varience_portfolio │ │ Feature: Class: Total: │ │ Positive variable 1 150 │ │ Total variables 1 150 │ │ Objective - 1 │ │ Constraint 1 2 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Solve ────────────────────────────────────────────────────────────────────────╮ │ │ │ Method: exact Objective value │ │ Status: max │ │ optimal* 1.19 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Metric ───────────────────────────────────────────────────────────────────────╮ │ │ │ CPT (microseconds) 6.67e+04 │ │ CPT (hour:min:sec) 00:00:00 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯ ╭─ Decision ─────────────────────────────────────────────────────────────────────╮ │ │ │ x[0] = 2.7743755325232772e-11 │ │ x[1] = 2.8092100079904077e-11 │ │ x[2] = 2.844887876112531e-11 │ │ x[3] = 2.8814401303430977e-11 │ │ x[4] = 2.918899301047485e-11 │ │ x[5] = 2.957299551998146e-11 │ │ x[6] = 2.9966767842550535e-11 │ │ x[7] = 3.0370687480921945e-11 │ │ x[8] = 3.078515163671375e-11 │ │ x[9] = 3.1210578513200454e-11 │ │ x[10] = 3.164740872293794e-11 │ │ x[11] = 3.209610681028805e-11 │ │ x[12] = 3.255716289959618e-11 │ │ x[13] = 3.3031094482143494e-11 │ │ x[14] = 3.351844835467661e-11 │ │ x[15] = 3.401980272547408e-11 │ │ x[16] = 3.4535769504910366e-11 │ │ x[17] = 3.5066996799521295e-11 │ │ x[18] = 3.56141716311133e-11 │ │ x[19] = 3.6178022905073607e-11 │ │ x[20] = 3.675932465476433e-11 │ │ x[21] = 3.735889959258624e-11 │ │ x[22] = 3.797762300220891e-11 │ │ x[23] = 3.861642701047935e-11 │ │ x[24] = 3.927630528350411e-11 │ │ x[25] = 3.995831819627192e-11 │ │ x[26] = 4.066359853329355e-11 │ │ x[27] = 4.139335778434345e-11 │ │ x[28] = 4.214889310940747e-11 │ │ x[29] = 4.2931595057495157e-11 │ │ x[30] = 4.3742956136354445e-11 │ │ x[31] = 4.4584580344059104e-11 │ │ x[32] = 4.545819379186986e-11 │ │ x[33] = 4.636565656614686e-11 │ │ x[34] = 4.730897600198497e-11 │ │ x[35] = 4.8290321568244213e-11 │ │ x[36] = 4.9312041597312165e-11 │ │ x[37] = 5.0376682131595416e-11 │ │ x[38] = 5.1487008206677906e-11 │ │ x[39] = 5.264602794597013e-11 │ │ x[40] = 5.385701991049141e-11 │ │ x[41] = 5.512356422858367e-11 │ │ x[42] = 5.644957812869436e-11 │ │ x[43] = 5.78393566196926e-11 │ │ x[44] = 5.929761920905631e-11 │ │ x[45] = 6.08295637309449e-11 │ │ x[46] = 6.244092857699688e-11 │ │ x[47] = 6.413806490058805e-11 │ │ x[48] = 6.592802070660395e-11 │ │ x[49] = 6.781863917117108e-11 │ │ x[50] = 6.981867407616489e-11 │ │ x[51] = 7.19379259316092e-11 │ │ x[52] = 7.418740323464869e-11 │ │ x[53] = 7.657951443926595e-11 │ │ x[54] = 7.912829766535345e-11 │ │ x[55] = 8.184969707238052e-11 │ │ x[56] = 8.47618973118352e-11 │ │ x[57] = 8.788573077302652e-11 │ │ x[58] = 9.12451767466872e-11 │ │ x[59] = 9.48679775894747e-11 │ │ x[60] = 9.878640510238738e-11 │ │ x[61] = 1.0303822156654325e-10 │ │ x[62] = 1.0766789557802846e-10 │ │ x[63] = 1.1272815508182758e-10 │ │ x[64] = 1.1828199202308608e-10 │ │ x[65] = 1.244052798588321e-10 │ │ x[66] = 1.311902348645078e-10 │ │ x[67] = 1.3875005793442852e-10 │ │ x[68] = 1.47225257515537e-10 │ │ x[69] = 1.567924144120609e-10 │ │ x[70] = 1.6767657252930607e-10 │ │ x[71] = 1.801691484175006e-10 │ │ x[72] = 1.9465447734618066e-10 │ │ x[73] = 2.1165030742284027e-10 │ │ x[74] = 2.3187164592398663e-10 │ │ x[75] = 2.56335359142733e-10 │ │ x[76] = 2.8653937163816974e-10 │ │ x[77] = 3.2478591996278845e-10 │ │ x[78] = 3.7479878731145546e-10 │ │ x[79] = 4.4296927756312283e-10 │ │ x[80] = 5.409807300816127e-10 │ │ x[81] = 6.916461389651319e-10 │ │ x[82] = 9.492864940447693e-10 │ │ x[83] = 1.5969233347143176e-09 │ │ x[84] = 4.949433819670803e-09 │ │ x[85] = 0.0004968283335643696 │ │ x[86] = 0.0011764259492214814 │ │ x[87] = 0.0018405992939069312 │ │ x[88] = 0.0024897413618520515 │ │ x[89] = 0.0031243765482852014 │ │ x[90] = 0.003745023440277047 │ │ x[91] = 0.004352162904818471 │ │ x[92] = 0.004946242752043668 │ │ x[93] = 0.005527684632444081 │ │ x[94] = 0.0060968889078715114 │ │ x[95] = 0.0066542377734675326 │ │ x[96] = 0.00720009720815459 │ │ x[97] = 0.0077348182350865735 │ │ x[98] = 0.008258737820428677 │ │ x[99] = 0.008772179596906758 │ │ x[100] = 0.009275454502787923 │ │ x[101] = 0.009768861374959598 │ │ x[102] = 0.010252687510635725 │ │ x[103] = 0.010727209202439966 │ │ x[104] = 0.011192692248249404 │ │ x[105] = 0.01164939243629286 │ │ x[106] = 0.012097556006046329 │ │ x[107] = 0.012537420085608268 │ │ x[108] = 0.012969213106455858 │ │ x[109] = 0.013393155196571777 │ │ x[110] = 0.013809458553019398 │ │ x[111] = 0.014218327795040875 │ │ x[112] = 0.014619960298729076 │ │ x[113] = 0.015014546514330065 │ │ x[114] = 0.015402270267140387 │ │ x[115] = 0.015783309042941812 │ │ x[116] = 0.016157834258887647 │ │ x[117] = 0.016526011520642342 │ │ x[118] = 0.016888000866597877 │ │ x[119] = 0.01724395699989295 │ │ x[120] = 0.017594029508926672 │ │ x[121] = 0.017938363077046653 │ │ x[122] = 0.01827709768198845 │ │ x[123] = 0.01861036878567794 │ │ x[124] = 0.018938307514905547 │ │ x[125] = 0.019261040833404748 │ │ x[126] = 0.019578691705782456 │ │ x[127] = 0.019891379253773885 │ │ x[128] = 0.020199218905216167 │ │ x[129] = 0.020502322536145638 │ │ x[130] = 0.020800798606385958 │ │ x[131] = 0.02109475228896583 │ │ x[132] = 0.021384285593700505 │ │ x[133] = 0.021669497485234675 │ │ x[134] = 0.021950483995841062 │ │ x[135] = 0.02222733833323745 │ │ x[136] = 0.022500150983681487 │ │ x[137] = 0.022769009810585366 │ │ x[138] = 0.023034000148868413 │ │ x[139] = 0.02329520489526858 │ │ x[140] = 0.023552704594807523 │ │ x[141] = 0.023806577523602854 │ │ x[142] = 0.024056899768205388 │ │ x[143] = 0.02430374530163019 │ │ x[144] = 0.024547186056240292 │ │ x[145] = 0.02478729199363727 │ │ x[146] = 0.025024131171698008 │ │ x[147] = 0.0252577698088989 │ │ x[148] = 0.025488272346048852 │ │ x[149] = 0.025715701381297428 │ │ │ ╰────────────────────────────────────────────────────────────────────────────────╯
- Note : Implementing this example at least requires installing the
-
Documentation
Citation
To cite or give credit to FelooPy in publications, projects, presentations, web pages, blog posts, etc. please use the following entries:
-
LaTeX:
@software{ktafakkori2022Sep, author = {Keivan Tafakkori}, title = {{FelooPy: An integrated optimization environment for AutoOR in Python}}, year = {2022}, month = sep, publisher = {GitHub}, url = {https://github.com/ktafakkori/feloopy/} }
-
APA:
Tafakkori, K. (2022). FelooPy: An integrated optimization environment for AutoOR in Python [Python Library]. Retrieved from https://github.com/ktafakkori/feloopy (Original work published September 2022). -
In-text:
- Note 1: Please write the version you used, the latest is v0.2.8.
- Note 2: Using secondary interfaces or solvers might also require a citation to their projects.
FelooPy (v0.2.8) was used in conjunction with [interface x] (v0.0.0) (except
feloopyitself) as the optimization interface and [solver y] (v0.0.0) as the optimization solver.
License
Copyright K. Tafakkori, 2022-2024
See the LICENSE file for copyright information.
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
feloopy-0.2.8.tar.gz
(107.5 kB
view details)
Built Distribution
feloopy-0.2.8-py3-none-any.whl
(161.2 kB
view details)
File details
Details for the file feloopy-0.2.8.tar.gz.
File metadata
- Download URL: feloopy-0.2.8.tar.gz
- Upload date:
- Size: 107.5 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 | 7e442a9b7afd8a32344aebdcef44fe03640f3da72d962578ea664a33849f1555 |
|
| MD5 | 30443d41ccd943b716a37dcdda920592 |
|
| BLAKE2b-256 | 1c5da182789743c09f4a91805148f37700989098930a00bc5c4a1b64c0898e06 |
File details
Details for the file feloopy-0.2.8-py3-none-any.whl.
File metadata
- Download URL: feloopy-0.2.8-py3-none-any.whl
- Upload date:
- Size: 161.2 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.6
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 | 2adf39e2c67c75e7211216ab360cc8a5c6925602cb5db382ce0cbf0a02986b36 |
|
| MD5 | d65a46c5ad74f740488d1a3bbea9e40d |
|
| BLAKE2b-256 | cb3f3df62f4c655037c4cc14d9c37c65c1c3423224f18abd591223ca8fe692cf |
