Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

float64 issue

See original GitHub issue

At the moment, HB only works with float32. You must cast float64 to float32 for correct results. (You will get an error with the gemm algorithm). We need to fix this.

Ex: in scikit-learn-random-forest-example.ipynb we must cast X as follows: X = X[0:nrows].astype('|f4')

Issue Analytics

State:
Created 3 years ago
Comments:13 (7 by maintainers)

Top GitHub Comments

1reaction

KranthiGVcommented, Jun 6, 2020

Okay, I’ll do that.

0reactions

interesaaatcommented, Jul 9, 2020

Ok that was fast! 😃 Don’t worry, I was just checking. These times are hard for everyone.

Top Results From Across the Web

python float64 type conversion issue with pandas

The problem is that a float64 a mantisse of 53 bits which can represent 15 or 16 decimal digits (ref). That means that...

m1 mac has different behavior when convert float64 to uint64

The issue is related to precision loss during hardware conversion from float64 to int64 . It looks like Apple M1 has slightly different...

Inaccurate float32 and float64: how to avoid the trap in Go ...

Photo by Luis Villasmil on Unsplash. Hi everyone! Let's talk about floats today. So, let's take a look on that simple code: var...

Float64 | Apple Developer Documentation

Float64. A 64-bit floating point type. iOS 8.0+ iPadOS 8.0+ macOS 10.10+ Mac Catalyst 13.0+ tvOS 9.0+ watchOS 2.0+. Declaration. typealias Float64 =...

Issue with PointType update from Float32 to Float64 - PI Square

Issue with PointType update from Float32 to Float64. Hi Team,. We have requirement to store more than 7 digits data in PI Tags....