How to contribute to openfhe-numpy?

OpenFHE-NumPy
1

Hello all, I think I have found a bug in the openfhe-numpy library that errors out for scalar subtraction. Here is the minimal script to reproduce this error:

import numpy as np

from openfhe import *

import openfhe_numpy as onp




# Cryptographic setup

mult_depth = 4

params = CCParamsCKKSRNS()




cc = GenCryptoContext(params)

params.SetMultiplicativeDepth(mult_depth)

cc.Enable(PKESchemeFeature.PKE)

cc.Enable(PKESchemeFeature.LEVELEDSHE)

cc.Enable(PKESchemeFeature.ADVANCEDSHE)




keys = cc.KeyGen()




ring_dim = cc.GetRingDimension()

batch_size = ring_dim // 2

print(f"\nCKKS ring dimension: {ring_dim}")

print(f"Available slots:    {batch_size}")




# Sample input vectors

vector = [1.0, 2.0, 3.0, 4.0, 5.0]




ct_v = onp.array(

cc=cc,

data=vector,

batch_size=batch_size,

order=onp.ROW_MAJOR,

mode="tile",

fhe_type="C",

public_key=keys.publicKey,

)




# Scalar value

scalar = 2.0




print(f"Vector: {vector}")

print(f"Scalar: {scalar}")




# Add scalar to vector

ct_v_add = ct_v + scalar

res_add = ct_v_add.decrypt(keys.secretKey, unpack_type="original")

print(f"Result (add): {res_add}")




# Subtract scalar from vector

ct_v_sub = ct_v + (-scalar)

res_sub_true = ct_v_sub.decrypt(keys.secretKey, unpack_type="original")

print(f"Result (sub true): {res_sub_true}")




# Subtract scalar from vector

ct_v_sub = ct_v - scalar # errors out

res_sub = ct_v_sub.decrypt(keys.secretKey, unpack_type="original")

print(f"Result (sub): {res_sub}")

I have prepared a fix for this in my local fork. But I am unable to create a PR on the parent repo.

Let me know if there is any other way to contribute.

2

Thank you. I will reply to you directly. The high-level idea is that we can add your github account to the repo so you could contribute to the openfhe-numpy project.

3

Hello Yuriy. My github username is mayank64ce.