Skip to content

Commit b157c19

Browse files
author
The ml_dtypes Authors
committed
Merge pull request #160 from jakevdp:fix-readme
PiperOrigin-RevId: 655150786
2 parents f739b2f + 04ab286 commit b157c19

File tree

1 file changed

+7
-7
lines changed

1 file changed

+7
-7
lines changed

README.md

Lines changed: 7 additions & 7 deletions
Original file line numberDiff line numberDiff line change
@@ -113,13 +113,13 @@ This type has the following characteristics:
113113
2 and 4-bit integer types, where each element is represented unpacked (i.e.,
114114
padded up to a byte in memory).
115115

116-
NumPy does not support types smaller than a single byte. For example, the
117-
distance between adjacent elements in an array (`.strides`) is expressed in
118-
bytes. Relaxing this restriction would be a considerable engineering project.
119-
These types therefore use an unpacked representation, where
120-
each element of the array is padded up to a byte in memory. The lower four bits
121-
of each byte contain the representation of the number, whereas the upper four
122-
bits are ignored.
116+
NumPy does not support types smaller than a single byte: for example, the
117+
distance between adjacent elements in an array (`.strides`) is expressed as
118+
an integer number of bytes. Relaxing this restriction would be a considerable
119+
engineering project. These types therefore use an unpacked representation, where
120+
each element of the array is padded up to a byte in memory. The lower two or four
121+
bits of each byte contain the representation of the number, whereas the remaining
122+
upper bits are ignored.
123123

124124
## Quirks of low-precision Arithmetic
125125

0 commit comments

Comments
 (0)