VOCS data structure¶

Variables, Objectives, Constraints, and other Settings (VOCS) helps define our optimization problems.

In [1]:

from xopt.vocs import VOCS
from xopt.vocs import form_objective_data
import pandas as pd
import numpy as np

from xopt.vocs import VOCS from xopt.vocs import form_objective_data import pandas as pd import numpy as np

In [2]:

Y = """
variables:
  a: [0, 1e3] # Note that 1e3 usually parses as a str with YAML. 
  b: [-1, 1]
objectives:
  c: maximize
  d: minimize 
constraints:
  e: ['Less_than', 2]
  f: ['greater_than', 0]
constants:
  g: 1234

"""

vocs = VOCS.from_yaml(Y)
vocs

Y = """ variables: a: [0, 1e3] # Note that 1e3 usually parses as a str with YAML. b: [-1, 1] objectives: c: maximize d: minimize constraints: e: ['Less_than', 2] f: ['greater_than', 0] constants: g: 1234 """ vocs = VOCS.from_yaml(Y) vocs

Out[2]:

VOCS(variables={'a': [0.0, 1000.0], 'b': [-1.0, 1.0]}, constraints={'e': ['LESS_THAN', 2.0], 'f': ['GREATER_THAN', 0.0]}, objectives={'c': 'MAXIMIZE', 'd': 'MINIMIZE'}, constants={'g': 1234}, observables=[])

In [3]:

# as dict
dict(vocs)

# as dict dict(vocs)

Out[3]:

{'variables': {'a': [0.0, 1000.0], 'b': [-1.0, 1.0]},
 'constraints': {'e': ['LESS_THAN', 2.0], 'f': ['GREATER_THAN', 0.0]},
 'objectives': {'c': 'MAXIMIZE', 'd': 'MINIMIZE'},
 'constants': {'g': 1234},
 'observables': []}

In [4]:

#  re-parse dict
vocs2 = VOCS.from_dict(dict(vocs))

# re-parse dict vocs2 = VOCS.from_dict(dict(vocs))

In [5]:

# Check that these are the same
vocs2 == vocs

# Check that these are the same vocs2 == vocs

Out[5]:

True

In [6]:

# This replaces the old vocs["variables"]
getattr(vocs, "variables")

# This replaces the old vocs["variables"] getattr(vocs, "variables")

Out[6]:

{'a': [0.0, 1000.0], 'b': [-1.0, 1.0]}

In [7]:

vocs.objectives["c"] == "MAXIMIZE"

vocs.objectives["c"] == "MAXIMIZE"

Out[7]:

True

In [8]:

# json
vocs.to_json()

# json vocs.to_json()

Out[8]:

'{"variables":{"a":[0.0,1000.0],"b":[-1.0,1.0]},"constraints":{"e":["LESS_THAN",2.0],"f":["GREATER_THAN",0.0]},"objectives":{"c":"MAXIMIZE","d":"MINIMIZE"},"constants":{"g":1234},"observables":[]}'

Objective Evaluation¶

In [9]:

data = pd.DataFrame(vocs.random_inputs(10))
# Add some outputs
data["c"] = data["a"] + data["b"]
data["d"] = data["a"] - data["b"]
data["e"] = data["a"] * 2 + data["b"] * 2
data["f"] = data["a"] * 2 - data["b"] * 2
data.index = np.arange(len(data)) + 5  # custom index
data

data = pd.DataFrame(vocs.random_inputs(10)) # Add some outputs data["c"] = data["a"] + data["b"] data["d"] = data["a"] - data["b"] data["e"] = data["a"] * 2 + data["b"] * 2 data["f"] = data["a"] * 2 - data["b"] * 2 data.index = np.arange(len(data)) + 5 # custom index data

Out[9]:

	a	b	g	c	d	e	f
5	429.279251	-0.653713	1234	428.625538	429.932964	857.251076	859.865928
6	902.089175	-0.534293	1234	901.554882	902.623468	1803.109765	1805.246937
7	861.252901	-0.802566	1234	860.450334	862.055467	1720.900669	1724.110935
8	146.984367	0.008667	1234	146.993034	146.975700	293.986068	293.951400
9	932.937804	0.450766	1234	933.388570	932.487037	1866.777140	1864.974074
10	591.789090	0.333188	1234	592.122278	591.455902	1184.244556	1182.911805
11	923.904891	0.300432	1234	924.205323	923.604460	1848.410646	1847.208919
12	109.985820	-0.546218	1234	109.439602	110.532038	218.879203	221.064075
13	11.928810	0.549522	1234	12.478332	11.379288	24.956665	22.758576
14	43.263545	-0.511541	1234	42.752004	43.775086	85.504008	87.550172

In [10]:

vocs.objectives

vocs.objectives

Out[10]:

{'c': 'MAXIMIZE', 'd': 'MINIMIZE'}

In [11]:

# These are in standard form for minimization
form_objective_data(vocs.objectives, data)

# These are in standard form for minimization form_objective_data(vocs.objectives, data)

Out[11]:

	objective_c	objective_d
5	-428.625538	429.932964
6	-901.554882	902.623468
7	-860.450334	862.055467
8	-146.993034	146.975700
9	-933.388570	932.487037
10	-592.122278	591.455902
11	-924.205323	923.604460
12	-109.439602	110.532038
13	-12.478332	11.379288
14	-42.752004	43.775086

In [12]:

# This is also available as a method
vocs.objective_data(data)

# This is also available as a method vocs.objective_data(data)

Out[12]:

	objective_c	objective_d
5	-428.625538	429.932964
6	-901.554882	902.623468
7	-860.450334	862.055467
8	-146.993034	146.975700
9	-933.388570	932.487037
10	-592.122278	591.455902
11	-924.205323	923.604460
12	-109.439602	110.532038
13	-12.478332	11.379288
14	-42.752004	43.775086

In [13]:

# use the to_numpy() method to convert for low level use.
vocs.objective_data(data).to_numpy()

# use the to_numpy() method to convert for low level use. vocs.objective_data(data).to_numpy()

Out[13]:

array([[-428.625538  ,  429.932964  ],
       [-901.55488237,  902.62346842],
       [-860.45033436,  862.0554673 ],
       [-146.99303387,  146.97569992],
       [-933.38857003,  932.48703714],
       [-592.12227789,  591.4559025 ],
       [-924.20532317,  923.60445957],
       [-109.43960163,  110.53203774],
       [ -12.47833241,   11.37928801],
       [ -42.75200409,   43.77508585]])

In [14]:

vocs.constraint_data(data)

vocs.constraint_data(data)

Out[14]:

	constraint_e	constraint_f
5	855.251076	-859.865928
6	1801.109765	-1805.246937
7	1718.900669	-1724.110935
8	291.986068	-293.951400
9	1864.777140	-1864.974074
10	1182.244556	-1182.911805
11	1846.410646	-1847.208919
12	216.879203	-221.064075
13	22.956665	-22.758576
14	83.504008	-87.550172

In [15]:

vocs.feasibility_data(data)

vocs.feasibility_data(data)

Out[15]:

	feasible_e	feasible_f	feasible
5	False	True	False
6	False	True	False
7	False	True	False
8	False	True	False
9	False	True	False
10	False	True	False
11	False	True	False
12	False	True	False
13	False	True	False
14	False	True	False

In [16]:

# normalize inputs to unit domain [0,1]
normed_data = vocs.normalize_inputs(data)
normed_data

# normalize inputs to unit domain [0,1] normed_data = vocs.normalize_inputs(data) normed_data

Out[16]:

	a	b
5	0.429279	0.173144
6	0.902089	0.232853
7	0.861253	0.098717
8	0.146984	0.504333
9	0.932938	0.725383
10	0.591789	0.666594
11	0.923905	0.650216
12	0.109986	0.226891
13	0.011929	0.774761
14	0.043264	0.244230

In [17]:

# and denormalize
vocs.denormalize_inputs(normed_data)

# and denormalize vocs.denormalize_inputs(normed_data)

Out[17]:

	a	b
5	429.279251	-0.653713
6	902.089175	-0.534293
7	861.252901	-0.802566
8	146.984367	0.008667
9	932.937804	0.450766
10	591.789090	0.333188
11	923.904891	0.300432
12	109.985820	-0.546218
13	11.928810	0.549522
14	43.263545	-0.511541

Error handling¶

In [18]:

Y = """
variables:
  a: [0, 1e3] # Note that 1e3 usually parses as a str with YAML. 
  b: [-1, 1]
objectives:
  c: maximize
  d: minimize 
constraints:
  e: ['Less_than', 2]
  f: ['greater_than', 0]
constants:
  g: 1234

"""

vocs = VOCS.from_yaml(Y)

Y = """ variables: a: [0, 1e3] # Note that 1e3 usually parses as a str with YAML. b: [-1, 1] objectives: c: maximize d: minimize constraints: e: ['Less_than', 2] f: ['greater_than', 0] constants: g: 1234 """ vocs = VOCS.from_yaml(Y)

In [19]:

d = {"a": [1, 2, 3]}

df = pd.DataFrame(d)
df2 = pd.DataFrame(df).copy()

df2["b"] = np.nan
df2["b"] - 1

d = {"a": [1, 2, 3]} df = pd.DataFrame(d) df2 = pd.DataFrame(df).copy() df2["b"] = np.nan df2["b"] - 1

Out[19]:

0   NaN
1   NaN
2   NaN
Name: b, dtype: float64

In [20]:

data["a"] = np.nan

data["a"] = np.nan

In [21]:

a = 2


def f(x=a):
    return x


a = 99
f()

a = 2 def f(x=a): return x a = 99 f()

Out[21]:

2

In [22]:

pd.DataFrame(6e66, index=[1, 2, 3], columns=["A"])

pd.DataFrame(6e66, index=[1, 2, 3], columns=["A"])

Out[22]:

	A
1	6.000000e+66
2	6.000000e+66
3	6.000000e+66

In [23]:

# These are in standard form for minimization

data = pd.DataFrame({"c": [1, 2, 3, 4]}, index=[9, 3, 4, 5])

form_objective_data(vocs.objectives, data)

# These are in standard form for minimization data = pd.DataFrame({"c": [1, 2, 3, 4]}, index=[9, 3, 4, 5]) form_objective_data(vocs.objectives, data)

Out[23]:

	objective_c	objective_d
9	-1.0	inf
3	-2.0	inf
4	-3.0	inf
5	-4.0	inf