Getting a large QUBO matrix (1,000 rows and columns) from Excel into Python/IDE

Tim B

December 23, 2020 14:45

Is there a way to bring a large QUBO matrix, with say 1,000 rows and columns, from Excel (i.e. values in cells) into Python code used by the Leap IDE Workspace? (If it was a small QUBO matrix then it is straightforward to use text to create the QUBO matrix in the Python code.)

Comments

1 comment

Tanvi M (Report)
December 30, 2020 01:30 (Edited December 30, 2020 01:33)
Hi Tim,

The functionality to convert a NumPy array to a binary quadratic model is available in Ocean SDK. Here is a list of all the conversions to and from other formats that are supported: https://docs.ocean.dwavesys.com/projects/dimod/en/latest/reference/bqm/binary_quadratic_model.html#converting-to-and-from-other-formats

So, in order to import data from an excel worksheet, save it to a csv file. This csv file can be loaded as a NumPy array by using numpy.loadtxt().

Here is an example:
```
import numpy as np
import dimod

f = open("Q_excel.csv")
Q = np.loadtxt(f,delimiter = ",")

bqm = dimod.BinaryQuadraticModel.from_numpy_matrix(Q)
```
For very large problems, dimod.AdjVectorBQM.from_numpy_matrix is recommended over dimod.BinaryQuadraticModel.from_numpy_matrix.

Please let us know if you have any further questions.
0

Comment actions Permalink

Please sign in to leave a comment.

Didn't find what you were looking for?

New post