-
https://labs.quansight.org/blog/numpy-quaddtype-blog
When this stabilizes, is this a candidate to replace bundled libqd? 🤔
-
I tried to convert the code from double precision to quadruple precision, and the code compiles well. However, while execution, it fails to read nets3 and nets4 files. Is there a proper way to read/wr…
-
LLVM might have some support for quadruple precision, but we need to figure out how to use it (https://github.com/lfortran/lfortran/pull/3037) and likely it will only be a partial support anyway or ha…
-
### Reproducing code example:
```python
import numpy as np
np.exp(8j * np.arctan(np.array(1, dtype=np.float16))).imag
np.exp…
-
From the spec:
```
print(3.14 ==== 3.14)! //true
```
This should be `false` since those are two independent literals.
However the following makes sense because the programmer probably reall…
-
This is the definition of `integer_decode`:
```rust
fn integer_decode(self) -> (u64, i16, i8)
```
The 64 bits allocated for the mantissa is too small, 112 bits would be required to store the m…
-
### Discussed in https://github.com/JuliaInterop/Clang.jl/discussions/508
Originally posted by **amontoison** August 30, 2024
Some Fortran libraries, like `CUTEst`, can be easily compiled in q…
-
The quadruple library is called `libooptools-quad.a` but configure checks only for `libooptools.a`. Do you think it makes sense to check for both?
-
Some hardware doesn't support 32-bit precision. We should emulate higher precision, where necessary.
- [Double Precision in OpenGL and WebGL](http://blog.hvidtfeldts.net/index.php/2012/07/double-preci…
-
I suggest declaring variables as `real(kind=wp)` instead of just `real`, so that the codes can be run in double or even quadruple precision if desired, by changing the value of the parameter `wp`, as …