Dynex SDK

Module contents

Redistributions of source code must retain the above copyright notice, this list of
conditions and the following disclaimer.
Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
Neither the name of the copyright holder nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS “AS IS” AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

class dynex.BQM(bqm, relabel=True, logging=False, formula=2)[source]

Bases: object

Creates a model, which can be used by the sampler based on a Binary Quadratic Model (BQM) problem. The Dynex sampler needs a “model” object for sampling. Based on the problem formulation, multiple model classes are supported.

Parameters

bqm

The BQM to be used for this model (class:dimod.BinaryQuadraticModel)

relabel

Defines if the BQM’s variable names should be relabeled (bool)

logging

True to show model creation information, False to silence outputs (bool)

Returns

class:dynex.model

Example

Dimod’s dimod.binary.BinaryQuadraticModel (BQM) contains linear and quadratic biases for problems formulated as binary quadratic models as well as additional information such as variable labels and offset.

bqm = dimod.BinaryQuadraticModel({'x1': 1.0, 'x2': -1.5, 'x3': 2.0},
                         {('x1', 'x2'): 1.0, ('x2', 'x3'): -2.0},
                         0.0, dimod.BINARY)
model = dynex.BQM(bqm)

class dynex.CQM(cqm, relabel=True, logging=False, formula=2)[source]

Bases: object

Creates a model, which can be used by the sampler based on a Constraint Quadratic Model (CQM) problem. The Dynex sampler needs a “model” object for sampling. Based on the problem formulation, multiple model classes are supported.

Parameters

cqm

The BQM to be used for this model (class:dimod.ConstraintQuadraticModel)

relabel

Defines if the BQM’s variable names should be relabeled (bool)

logging

True to show model creation information, False to silence outputs (bool)

Returns

class:dynex.model

Example

Dimod’s dimod.ConstrainedQuadraticModel (CQM) contains linear and quadratic biases for problems formulated as constrained quadratic models as well as additional information such as variable labels, offsets, and equality and inequality constraints.

num_widget_a = dimod.Integer('num_widget_a', upper_bound=7)
num_widget_b = dimod.Integer('num_widget_b', upper_bound=3)
cqm = dimod.ConstrainedQuadraticModel()
cqm.set_objective(-3 * num_widget_a - 4 * num_widget_b)
cqm.add_constraint(num_widget_a + num_widget_b <= 5, label='total widgets')
model = dynex.CQM(cqm)

class dynex.DynexSampler(model, logging=True, mainnet=True, description='Dynex SDK Job', test=False, bnb=True)[source]

Bases: object

Initialises the sampler object given a model.

Parameters

logging

Defines if the sampling process should be quiet with no terminal output (FALSE) or if process updates are to be shown (bool)
mainnet

Defines if the mainnet (Dynex platform sampling) or the testnet (local sampling) is being used for sampling (bool)
description

Defines the description for the sampling, which is shown in Dynex job dashboards as well as in the market place (string)

Returns

class:dynex.samper

Example

sampler = dynex.DynexSampler(model)

sample(num_reads=32, annealing_time=10, clones=1, switchfraction=0.0, alpha=20, beta=20, gamma=1, delta=1, epsilon=1, zeta=1, minimum_stepsize=6e-08, debugging=False, block_fee=0)[source]

The main sampling function:

Parameters

num_reads

Defines the number of parallel samples to be performed (int value in the range of [32, MAX_NUM_READS] as defined in your license)
annealing_time

Defines the number of integration steps for the sampling. Models are being converted into neuromorphic circuits, which are then simulated with ODE integration by the participating workers (int value in the range of [1, MAX_ANNEALING_TIME] as defined in your license)
clones

Defines the number of clones being used for sampling. Default value is 1 which means that no clones are being sampled. Especially when all requested num_reads will fit on one worker, it is desired to also retrieve the optimum ground states found from more than just one worker. The number of clones runs the sampler for n clones in parallel and aggregates the samples. This ensures a broader spectrum of retrieved samples. Please note, it the number of clones is set higher than the number of available threads on your local machine, then the number of clones processed in parallel is being processed in batches. Clone sampling is only available when sampling on the mainnet. (integer value in the range of [1,128])
switchfraction

Defines the percentage of variables which are replaced by random values during warm start samplings (double in the range of [0.0, 1.0])
alpha

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
beta

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
gamma

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
delta

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
epsilon

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
zeta

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
minimum_stepsize

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is performig adaptive stepsizes for each ODE integration forward Euler step. This value defines the smallest step size for a single adaptive integration step (double value in the range of [0.0000000000000001, 1.0])
debugging

Only applicable for test-net sampling. Defines if the sampling process should be quiet with no terminal output (FALSE) or if process updates are to be shown (TRUE) (bool)
bnb

Use alternative branch-and-bound sampling when in mainnet=False (bool)
block_fee

Computing jobs on the Dynex platform are being prioritised by the block fee which is being offered for computation. If this parameter is not specified, the current average block fee on the platform is being charged. To set an individual block fee for the sampling, specify this parameter, which is the amount of DNX in nanoDNX (1 DNX = 1,000,000,000 nanoDNX)

Returns

Returns a dimod sampleset object class:dimod.sampleset

Example

import dynex
import dimod

# Define the QUBU problem:
bqmodel = dimod.BinaryQuadraticModel({0: -1, 1: -1}, {(0, 1): 2}, 0.0, dimod.BINARY)

# Sample the problem:
model = dynex.BQM(bqmodel)
sampler = dynex.DynexSampler(model)
sampleset = sampler.sample(num_reads=32, annealing_time = 100)

# Output the result:
print(sampleset)

╭────────────┬───────────┬───────────┬─────────┬─────┬─────────┬───────┬─────┬──────────┬──────────╮
│   DYNEXJOB │   ELAPSED │   WORKERS │   CHIPS │   ✔ │   STEPS │   LOC │   ✔ │   ENERGY │        ✔ │
├────────────┼───────────┼───────────┼─────────┼─────┼─────────┼───────┼─────┼──────────┼──────────┤
│       3617 │      0.07 │         1 │       0 │  32 │     100 │     0 │   1 │        0 │ 10000.00 │
╰────────────┴───────────┴───────────┴─────────┴─────┴─────────┴───────┴─────┴──────────┴──────────╯
╭─────────────────────────────┬───────────┬─────────┬───────┬──────────┬───────────┬───────────────┬──────────╮
│                      WORKER │   VERSION │   CHIPS │   LOC │   ENERGY │   RUNTIME │   LAST UPDATE │   STATUS │
├─────────────────────────────┼───────────┼─────────┼───────┼──────────┼───────────┼───────────────┼──────────┤
│ *** WAITING FOR WORKERS *** │           │         │       │          │           │               │          │
╰─────────────────────────────┴───────────┴─────────┴───────┴──────────┴───────────┴───────────────┴──────────╯
[DYNEX] FINISHED READ AFTER 0.07 SECONDS
[DYNEX] PARSING 1 VOLTAGE ASSIGNMENT FILES...
progress: 100%
1/1 [00:05<00:00, 5.14s/it]
[DYNEX] SAMPLESET LOADED
[DYNEX] MALLOB: JOB UPDATED: 3617 STATUS: 2
   0  1 energy num_oc.
0  0  1   -1.0       1
['BINARY', 1 rows, 1 samples, 2 variables]

class dynex.SAT(clauses, logging=False)[source]

Bases: object

Creates a model, which can be used by the sampler based on a SAT problem. The Dynex sampler needs a “model” object for sampling. Based on the problem formulation, multiple model classes are supported.

Parameters

clauses

List of sat caluses for this model (list)

logging

True to show model creation information, False to silence outputs (bool)

Returns

class:dynex.model

Example

Dimod’s dimod.binary.BinaryQuadraticModel (BQM) contains linear and quadratic biases for problems formulated as binary quadratic models as well as additional information such as variable labels and offset.

clauses = [[1, -2, 3], [-1, 4, 5], [6, 7, -8], [-9, -10, 11], [12, 13, -14],
   [-1, 15, -16], [17, -18, 19], [-20, 2, 3], [4, -5, 6], [-7, 8, 9],
   [10, 11, -12], [-13, -14, 15], [16, 17, -18], [-19, 20, 1], [2, -3, 4],
   [-5, 6, 7], [8, 9, -10], [-11, -12, 13], [14, 15, -16], [-17, 18, 19]]
model =  dynex.SAT(clauses)

dynex.account_status()[source]

Shows the status of the Dynex SDK account as well as subscription information:

ACCOUNT: <YOUR ACCOUNT IDENTIFICATION>
API SUCCESSFULLY CONNECTED TO DYNEX
-----------------------------------
MAXIMUM NUM_READS: 100,000
MAXIMUM ANNEALING_TIME: 10,000
MAXIMUM JOB DURATION: 60 MINUTES
COMPUTE:
CURRENT AVG BLOCK FEE: 31.250005004 DNX
USAGE:
AVAILABLE BALANCE: 90.0 DNX
USAGE TOTAL: 0.0 DNX

dynex.estimate_costs(model, num_reads, block_fee=0)[source]

Dynex API call to estimate costs for a given job by analysing its complexity, given the number of Dynex chips (num_reads) to be used by calculating its network participation.

Parameters

x

model

Model to compute
num_reads

Number of Dynex chips to use
block_fee

(optional) block fee to override current network average fee

Returns

price_per_block

Effective price in nanoDNX (/1e9 for DNX) per block
price_per_minute

Effective price in nanoDNX (/1e9 for DNX) per minute
job_type

1 = SAT, 3 = MAXSAT, 5 = ISING/QUBO

Example

model = dynex.BQM(bqm);
dynex.estimate_costs(model, num_reads=10000);
[DYNEX] AVERAGE BLOCK FEE: 282.59 DNX
[DYNEX] SUBMITTING COMPUTE FILE FOR COST ESTIMATION...
[DYNEX] COST OF COMPUTE: 0.537993485 DNX PER BLOCK
[DYNEX] COST OF COMPUTE: 0.268996742 DNX PER MINUTE

dynex.sample(bqm, logging=True, formula=2, mainnet=False, description='Dynex SDK Job', test=False, bnb=True, num_reads=32, annealing_time=10, clones=1, switchfraction=0.0, alpha=20, beta=20, gamma=1, delta=1, epsilon=1, zeta=1, minimum_stepsize=6e-08, debugging=False, block_fee=0)[source]

Samples a Binary Quadratic Model (bqm).

Parameters

bqm

Binary quadratic model to sample
logging

Defines if the sampling process should be quiet with no terminal output (FALSE) or if process updates are to be shown (bool)
mainnet

Defines if the mainnet (Dynex platform sampling) or the testnet (local sampling) is being used for sampling (bool)
description

Defines the description for the sampling, which is shown in Dynex job dashboards as well as in the market place (string)
num_reads

Defines the number of parallel samples to be performed (int value in the range of [32, MAX_NUM_READS] as defined in your license)
annealing_time

Defines the number of integration steps for the sampling. Models are being converted into neuromorphic circuits, which are then simulated with ODE integration by the participating workers (int value in the range of [1, MAX_ANNEALING_TIME] as defined in your license)
clones

Defines the number of clones being used for sampling. Default value is 1 which means that no clones are being sampled. Especially when all requested num_reads will fit on one worker, it is desired to also retrieve the optimum ground states found from more than just one worker. The number of clones runs the sampler for n clones in parallel and aggregates the samples. This ensures a broader spectrum of retrieved samples. Please note, it the number of clones is set higher than the number of available threads on your local machine, then the number of clones processed in parallel is being processed in batches. Clone sampling is only available when sampling on the mainnet. (integer value in the range of [1,128])
switchfraction

Defines the percentage of variables which are replaced by random values during warm start samplings (double in the range of [0.0, 1.0])
alpha

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
beta

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
gamma

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
delta

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
epsilon

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
zeta

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
minimum_stepsize

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is performig adaptive stepsizes for each ODE integration forward Euler step. This value defines the smallest step size for a single adaptive integration step (double value in the range of [0.0000000000000001, 1.0])
debugging

Only applicable for test-net sampling. Defines if the sampling process should be quiet with no terminal output (FALSE) or if process updates are to be shown (TRUE) (bool)
bnb

Use alternative branch-and-bound sampling when in mainnet=False (bool)
block_fee

Computing jobs on the Dynex platform are being prioritised by the block fee which is being offered for computation. If this parameter is not specified, the current average block fee on the platform is being charged. To set an individual block fee for the sampling, specify this parameter, which is the amount of DNX in nanoDNX (1 DNX = 1,000,000,000 nanoDNX)

Returns

Returns a dimod sampleset object class:dimod.sampleset

Example

from pyqubo import Array
N = 15
K = 3
numbers = [4.8097315016016315, 4.325157567810298, 2.9877429101815127,
           3.199880179616316, 0.5787939511978596, 1.2520928214246918,
           2.262867466401502, 1.2300003067401255, 2.1601079352817925,
           3.63753899583021, 4.598232793833491, 2.6215815162575646,
           3.4227134835783364, 0.28254151584552023, 4.2548151473817075]

q = Array.create('q', N, 'BINARY')
H = sum(numbers[i] * q[i] for i in range(N)) + 5.0 * (sum(q) - K)**2
model = H.compile()
Q, offset = model.to_qubo(index_label=True)
bqm = dimod.BinaryQuadraticModel.from_qubo(Q, offset)
sampleset = dynex.sample(bqm, offset, formula=2, annealing_time=200, bnb=True)
print(sampleset)
   0  1  2  3  4  5  6  7  8  9 10 11 12 13 14   energy num_oc.
0  0  1  0  0  0  0  0  1  0  0  1  0  0  0  0 2.091336       1
['BINARY', 1 rows, 1 samples, 15 variables]

dynex.sample_ising(h, j, logging=True, formula=2, mainnet=False, description='Dynex SDK Job', test=False, bnb=True, num_reads=32, annealing_time=10, clones=1, switchfraction=0.0, alpha=20, beta=20, gamma=1, delta=1, epsilon=1, zeta=1, minimum_stepsize=6e-08, debugging=False, block_fee=0)[source]

Samples an Ising problem.

Parameters

h

Linear biases of the Ising problem
j

Quadratic biases of the Ising problem
logging

Defines if the sampling process should be quiet with no terminal output (FALSE) or if process updates are to be shown (bool)
mainnet

Defines if the mainnet (Dynex platform sampling) or the testnet (local sampling) is being used for sampling (bool)
description

Defines the description for the sampling, which is shown in Dynex job dashboards as well as in the market place (string)
num_reads

Defines the number of parallel samples to be performed (int value in the range of [32, MAX_NUM_READS] as defined in your license)
annealing_time

Defines the number of integration steps for the sampling. Models are being converted into neuromorphic circuits, which are then simulated with ODE integration by the participating workers (int value in the range of [1, MAX_ANNEALING_TIME] as defined in your license)
clones

Defines the number of clones being used for sampling. Default value is 1 which means that no clones are being sampled. Especially when all requested num_reads will fit on one worker, it is desired to also retrieve the optimum ground states found from more than just one worker. The number of clones runs the sampler for n clones in parallel and aggregates the samples. This ensures a broader spectrum of retrieved samples. Please note, it the number of clones is set higher than the number of available threads on your local machine, then the number of clones processed in parallel is being processed in batches. Clone sampling is only available when sampling on the mainnet. (integer value in the range of [1,128])
switchfraction

Defines the percentage of variables which are replaced by random values during warm start samplings (double in the range of [0.0, 1.0])
alpha

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
beta

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
gamma

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
delta

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
epsilon

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
zeta

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
minimum_stepsize

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is performig adaptive stepsizes for each ODE integration forward Euler step. This value defines the smallest step size for a single adaptive integration step (double value in the range of [0.0000000000000001, 1.0])
debugging

Only applicable for test-net sampling. Defines if the sampling process should be quiet with no terminal output (FALSE) or if process updates are to be shown (TRUE) (bool)
bnb

Use alternative branch-and-bound sampling when in mainnet=False (bool)
block_fee

Computing jobs on the Dynex platform are being prioritised by the block fee which is being offered for computation. If this parameter is not specified, the current average block fee on the platform is being charged. To set an individual block fee for the sampling, specify this parameter, which is the amount of DNX in nanoDNX (1 DNX = 1,000,000,000 nanoDNX)

Returns

Returns a dimod sampleset object class:dimod.sampleset

dynex.sample_qubo(Q, offset=0.0, logging=True, formula=2, mainnet=False, description='Dynex SDK Job', test=False, bnb=True, num_reads=32, annealing_time=10, clones=1, switchfraction=0.0, alpha=20, beta=20, gamma=1, delta=1, epsilon=1, zeta=1, minimum_stepsize=6e-08, debugging=False, block_fee=0)[source]

Samples a Qubo problem.

Parameters

Q

The Qubo problem
offset

The offset value of the Qubo problem
logging

Defines if the sampling process should be quiet with no terminal output (FALSE) or if process updates are to be shown (bool)
mainnet

Defines if the mainnet (Dynex platform sampling) or the testnet (local sampling) is being used for sampling (bool)
description

Defines the description for the sampling, which is shown in Dynex job dashboards as well as in the market place (string)
num_reads

Defines the number of parallel samples to be performed (int value in the range of [32, MAX_NUM_READS] as defined in your license)
annealing_time

Defines the number of integration steps for the sampling. Models are being converted into neuromorphic circuits, which are then simulated with ODE integration by the participating workers (int value in the range of [1, MAX_ANNEALING_TIME] as defined in your license)
clones

Defines the number of clones being used for sampling. Default value is 1 which means that no clones are being sampled. Especially when all requested num_reads will fit on one worker, it is desired to also retrieve the optimum ground states found from more than just one worker. The number of clones runs the sampler for n clones in parallel and aggregates the samples. This ensures a broader spectrum of retrieved samples. Please note, it the number of clones is set higher than the number of available threads on your local machine, then the number of clones processed in parallel is being processed in batches. Clone sampling is only available when sampling on the mainnet. (integer value in the range of [1,128])
switchfraction

Defines the percentage of variables which are replaced by random values during warm start samplings (double in the range of [0.0, 1.0])
alpha

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
beta

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
gamma

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
delta

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
epsilon

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
zeta

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is using automatic tuning of these parameters for the ODE integration. Setting values defines the upper bound for the automated parameter tuning (double value in the range of [0.00000001, 100.0] for alpha and beta, and [0.0 and 1.0] for gamma, delta and epsilon)
minimum_stepsize

The ODE integration of the QUBU/Ising or SAT model based neuromorphic circuits is performig adaptive stepsizes for each ODE integration forward Euler step. This value defines the smallest step size for a single adaptive integration step (double value in the range of [0.0000000000000001, 1.0])
debugging

Only applicable for test-net sampling. Defines if the sampling process should be quiet with no terminal output (FALSE) or if process updates are to be shown (TRUE) (bool)
bnb

Use alternative branch-and-bound sampling when in mainnet=False (bool)
block_fee

Computing jobs on the Dynex platform are being prioritised by the block fee which is being offered for computation. If this parameter is not specified, the current average block fee on the platform is being charged. To set an individual block fee for the sampling, specify this parameter, which is the amount of DNX in nanoDNX (1 DNX = 1,000,000,000 nanoDNX)

Returns

Returns a dimod sampleset object class:dimod.sampleset

Example

from pyqubo import Array
N = 15
K = 3
numbers = [4.8097315016016315, 4.325157567810298, 2.9877429101815127,
           3.199880179616316, 0.5787939511978596, 1.2520928214246918,
           2.262867466401502, 1.2300003067401255, 2.1601079352817925,
           3.63753899583021, 4.598232793833491, 2.6215815162575646,
           3.4227134835783364, 0.28254151584552023, 4.2548151473817075]

q = Array.create('q', N, 'BINARY')
H = sum(numbers[i] * q[i] for i in range(N)) + 5.0 * (sum(q) - K)**2
model = H.compile()
Q, offset = model.to_qubo(index_label=True)
sampleset = dynex.sample_qubo(Q, offset, formula=2, annealing_time=200, bnb=True)
print(sampleset)
   0  1  2  3  4  5  6  7  8  9 10 11 12 13 14   energy num_oc.
0  0  1  0  0  0  0  0  1  0  0  1  0  0  0  0 2.091336       1
['BINARY', 1 rows, 1 samples, 15 variables]

dynex.test()[source]: Performs test of the dynex.ini settings. Successful completion is required to start using the sampler.