Skip to content

Latest commit

 

History

History
96 lines (63 loc) · 2.75 KB

quickstart.rst

File metadata and controls

96 lines (63 loc) · 2.75 KB
Raw

Quickstart

GEFEST Framework quick start guide

How to install

pip install https://github.com/ITMO-NSS-team/GEFEST/archive/master.zip

How to design your own polygon in manual way

  • Step 1. Create loss function or simulator of the physical process.

Loss function for finding a polygon that seems like circle showed below.

import numpy as np
from gefest.core.geometry.geometry_2d import Geometry2D

geometry = Geometry2D()


def circle_loss(polygon):
   #calculating area and length of designed polygon via GEFEST methods
   area = geometry.get_square(polygon)
   length = geometry.get_length(polygon)

   if area == 0:
       return None
   #checking "area/length" ratio (equal 1 for circle)
   ratio = 4 * np.pi * area / length ** 2

   loss = 1 - ratio
   return loss
  • Step 2. Specify border coordinates of area where GEFEST will solve the task.

Put the Domain to Setup() class for creating a task variable.

from gefest.core.structure.domain import Domain
from gefest.core.opt.setup import Setup

border_coord = [(0, 0), (0, 300), (300, 300),
                (300, 0), (0, 0)]

domain = Domain(allowed_area=border_coord,

               # specify processing way
               geometry=geometry,

               # every designed polygon locates into Structure(),
               # these parameters determine number of polygons per Structure()
               max_poly_num=7,
               min_poly_num=1,

               # every designed polygon might сontain up to 20 points
               max_points_num=20,
               min_points_num=5,

               # designed polygons have closed borders
               is_closed=True)

task_setup = Setup(domain=domain)
  • Step 3. Create an optimized structure via optimize method.

The structure will contain number of polygons that previously specified in Domain

from gefest.core.opt.optimize import optimize

optimized_structure = optimize(task_setup=task_setup,
                               objective_function=circle_loss,

                               # Choosen population size and max number of generations
                               # for evolutionary optimization process
                               pop_size=100,
                               max_gens=220)
  • Step 4. Create visualization of designed structure.
from gefest.core.viz.struct_vizualizer import StructVizualizer

visualiser = StructVizualizer(task_setup.domain)
plt.figure(figsize=(7, 7))

info = {'fitness': circle_loss(optimized_structure.polygons[0]),
        'type': 'prediction'}
visualiser.plot_structure(optimized_structure, info)