RCDS

RCDS Optimization¶

In this example we demonstrate RCDS optimization.

In [1]:

# If you encounter the "Initializing libomp.dylib, but found libomp.dylib already initialized." error
# Please run this cell
import os

# If you encounter the "Initializing libomp.dylib, but found libomp.dylib already initialized." error # Please run this cell import os

In [2]:

import numpy as np
from xopt.generators.sequential import RCDSGenerator
from xopt.vocs import VOCS
from xopt.evaluator import Evaluator
from xopt import Xopt
from tqdm.auto import tqdm
import warnings


warnings.filterwarnings("ignore")

os.environ["KMP_DUPLICATE_LIB_OK"] = "True"

import numpy as np from xopt.generators.sequential import RCDSGenerator from xopt.vocs import VOCS from xopt.evaluator import Evaluator from xopt import Xopt from tqdm.auto import tqdm import warnings warnings.filterwarnings("ignore") os.environ["KMP_DUPLICATE_LIB_OK"] = "True"

In [3]:

# Reload packages automatically
%load_ext autoreload
%autoreload 2

# Reload packages automatically %load_ext autoreload %autoreload 2

RCDS test problem¶

This test problem is a 2-D quadratic function.

In [4]:

def f_test(input_dict):
    p = []
    for i in range(2):
        p.append(input_dict[f"p{i}"])

    obj = np.linalg.norm(p)
    outcome_dict = {"f": obj}

    return outcome_dict

def f_test(input_dict): p = [] for i in range(2): p.append(input_dict[f"p{i}"]) obj = np.linalg.norm(p) outcome_dict = {"f": obj} return outcome_dict

Run RCDS on the test problem (YAML method)¶

In [5]:

YAML = """
max_evaluations: 100
generator:
    name: rcds
    x0: null
    init_mat: null
    noise: 0.00001
    step: 0.01
evaluator:
    function: __main__.f_test
vocs:
    variables:
        p0: [0, 1]
        p1: [0, 1]
    objectives:
        f: MINIMIZE
"""

X = Xopt.from_yaml(YAML)
X

YAML = """ max_evaluations: 100 generator: name: rcds x0: null init_mat: null noise: 0.00001 step: 0.01 evaluator: function: __main__.f_test vocs: variables: p0: [0, 1] p1: [0, 1] objectives: f: MINIMIZE """ X = Xopt.from_yaml(YAML) X

Out[5]:

            Xopt
________________________________
Version: 0.1.dev1947+g7831d28.d20250426
Data size: 0
Config as YAML:
dump_file: null
evaluator:
  function: __main__.f_test
  function_kwargs: {}
  max_workers: 1
  vectorized: false
generator:
  init_mat: null
  is_active: false
  name: rcds
  noise: 1.0e-05
  step: 0.01
  supports_single_objective: true
max_evaluations: 100
serialize_inline: false
serialize_torch: false
strict: true
vocs:
  constants: {}
  constraints: {}
  objectives:
    f: MINIMIZE
  observables: []
  variables:
    p0:
    - 0.0
    - 1.0
    p1:
    - 0.0
    - 1.0

In [6]:

X.random_evaluate(1)
X.run()

X.random_evaluate(1) X.run()

Now you can go directly to the Visualization section and check out the results.

Run RCDS on the maximization test problem (YAML method)¶

In [7]:

YAML = """
max_evaluations: 400
generator:
    name: rcds
    x0: null
    init_mat: null
    noise: 1e-8
    step: 0.01
evaluator:
    function: __main__.f_test
vocs:
    variables:
        p0: [1, 3]
        p1: [-4, 2]
    objectives:
        f: MAXIMIZE
"""

X = Xopt.from_yaml(YAML)
X

YAML = """ max_evaluations: 400 generator: name: rcds x0: null init_mat: null noise: 1e-8 step: 0.01 evaluator: function: __main__.f_test vocs: variables: p0: [1, 3] p1: [-4, 2] objectives: f: MAXIMIZE """ X = Xopt.from_yaml(YAML) X

Out[7]:

            Xopt
________________________________
Version: 0.1.dev1947+g7831d28.d20250426
Data size: 0
Config as YAML:
dump_file: null
evaluator:
  function: __main__.f_test
  function_kwargs: {}
  max_workers: 1
  vectorized: false
generator:
  init_mat: null
  is_active: false
  name: rcds
  noise: 1.0e-08
  step: 0.01
  supports_single_objective: true
max_evaluations: 400
serialize_inline: false
serialize_torch: false
strict: true
vocs:
  constants: {}
  constraints: {}
  objectives:
    f: MAXIMIZE
  observables: []
  variables:
    p0:
    - 1.0
    - 3.0
    p1:
    - -4.0
    - 2.0

In [8]:

X.random_evaluate(1)
X.run()

X.random_evaluate(1) X.run()

Now you can go directly to the Visualization section and check out the results.

Run RCDS on the test problem (API method)¶

VOCS¶

We'll set the bounds for all the variables pi to [0, 1].

In [9]:

n_var = 2

n_var = 2

In [10]:

variables = {}
for i in range(n_var):
    variables[f"p{i}"] = [0, 1]

vocs = VOCS(
    variables=variables,
    objectives={"f": "MINIMIZE"},
)

variables = {} for i in range(n_var): variables[f"p{i}"] = [0, 1] vocs = VOCS( variables=variables, objectives={"f": "MINIMIZE"}, )

In [11]:

vocs

vocs

Out[11]:

VOCS(variables={'p0': [0.0, 1.0], 'p1': [0.0, 1.0]}, constraints={}, objectives={'f': 'MINIMIZE'}, constants={}, observables=[])

Evaluator¶

In [12]:

evaluator = Evaluator(function=f_test)

evaluator = Evaluator(function=f_test)

Generator¶

In [13]:

generator = RCDSGenerator(vocs=vocs)

generator = RCDSGenerator(vocs=vocs)

In [14]:

generator.model_dump()

generator.model_dump()

Out[14]:

{'supports_single_objective': True,
 'is_active': False,
 'init_mat': None,
 'noise': 1e-05,
 'step': 0.01}

Run the optimization¶

In [15]:

X = Xopt(vocs=vocs, evaluator=evaluator, generator=generator)

X = Xopt(vocs=vocs, evaluator=evaluator, generator=generator)

In [16]:

X.evaluate_data({"p0": 0.5, "p1": 0.5})

X.evaluate_data({"p0": 0.5, "p1": 0.5})

Out[16]:

	p0	p1	f	xopt_runtime	xopt_error
0	0.5	0.5	0.707107	0.000018	False

In [17]:

for i in tqdm(range(400)):
    X.step()

for i in tqdm(range(400)): X.step()

In [ ]:

Visualization¶

In [18]:

X.data.plot(y="f")

X.data.plot(y="f")

Out[18]:

<Axes: >

In [ ]: