Getting irregular output when setting "returnVectors" parameter in .detect differently

[bowenyi-pierre]

Hello,

I was using the .detect function in pycld2 on a Spanish sentence. Ideally, details[0][0] would be "SPANISH".

However, if I set returnVectors=False, details[0][0] was "ENGLISH." If I set returnVectors=False, details[0][0] was "SPANISH." Please see the attachment below for more details.

I am not sure what is going on here, and I think setting "returnVectors" parameter differently won't cause irregular detection of the most dominant language. Is this possibly a bug to fix? Thank you very much!

Best,
BW