commit ce83739f232a64c39f43d2e2b72f3af6e03635b1
parent 6f4c975bd4324e5c77e9ac1c83123c44bce8623f
Author: NunoSempere <nuno.sempere@protonmail.com>
Date: Sat, 24 Feb 2024 17:08:39 -0300
add well-tuned numpy python implementation
Diffstat:
2 files changed, 13 insertions(+), 3 deletions(-)
diff --git a/README.md b/README.md
@@ -29,8 +29,9 @@ The name of this repository is a pun on two meanings of "time to": "how much tim
| go | 32.70ms | 150 |
| Nim | 41.10ms | 84 |
| Lua (LuaJIT) | 68.80ms | 82 |
-| OCaml (flambda) | 185.50ms | 123 |
-| Squiggle (bun) | 384.00ms | 14* |
+| Python (well-tuned numpy) | 118ms | 123 |
+| OCaml (flambda) | 185ms | 123 |
+| Squiggle (bun) | 384ms | 14* |
| Javascript (node) | 0.423s | 69 |
| SquigglePy (v0.27) | 1.542s | 18* |
| R (3.6.1) | 4.494s | 49 |
@@ -96,7 +97,7 @@ In general, Squiggle tries to present a simple interface to the user, leading to
### Python and Squigglepy
-For the Python code, it's possible that the lack of speed is more a function of me not being as familiar with Python. It's also very possible that the code would run faster with [PyPy](https://doc.pypy.org).
+For the Python code, it's possible that the lack of speed is more a function of me not being as familiar with Python. It's also very possible that the code would run faster with [PyPy](https://doc.pypy.org). Update Feb/2024: Yeah, some well-tuned numpy code contributed by Jonathan Mann runs much faster.
In terms of complexity, SquigglePy seems to be between squiggle.c and the original squiggle. Like the original suqiggle, it also hides its stuff behind semi-magic wrappers, leading to e.g. ambiguities like around [correlated samples](https://git.nunosempere.com/personal/squiggle.c#correlated-samples) and generally having moving pieces that I don't understand. On the other hand, the code *is* short enough so that one person could read it in a few afternoons and roughly understand it. In terms of speed, SquigglePy seems slow.
diff --git a/time.txt b/time.txt
@@ -97,3 +97,12 @@ Time using 16 threads: 7.20ms
make time-linux
Running 100x and taking avg time: ./squiggle
Time using 16 threads: 32.70ms
+
+# Python (serialized numpy)
+
+time python3 samples-fast.py
+Mean result: 0.8848672124030412
+
+real 0m0.117s
+user 0m0.146s
+sys 0m0.298s
