Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/joekain/bmark
A benchmarking tool for Elixir
https://github.com/joekain/bmark
Last synced: about 2 months ago
JSON representation
A benchmarking tool for Elixir
- Host: GitHub
- URL: https://github.com/joekain/bmark
- Owner: joekain
- License: mit
- Created: 2015-02-15T22:46:14.000Z (over 9 years ago)
- Default Branch: master
- Last Pushed: 2018-03-11T03:29:21.000Z (over 6 years ago)
- Last Synced: 2024-04-01T04:47:20.389Z (3 months ago)
- Language: Elixir
- Homepage:
- Size: 57.6 KB
- Stars: 70
- Watchers: 3
- Forks: 3
- Open Issues: 5
-
Metadata Files:
- Readme: README.md
- Contributing: CONTRIBUTING.md
- License: LICENSE
Lists
- awesome-elixir - bmark - A benchmarking tool for Elixir. (Benchmarking)
- freaking_awesome_elixir - Elixir - A benchmarking tool for Elixir. (Benchmarking)
- awesome-elixir - bmark - A benchmarking tool for Elixir. (Benchmarking)
- fucking-awesome-elixir - bmark - A benchmarking tool for Elixir. (Benchmarking)
README
# Bmark
[](https://travis-ci.org/joekain/bmark) [](http://inch-ci.org/github/joekain/bmark) [](https://hex.pm/packages/bmark)
Bmark is a benchmarking tool for Elixir. It allows easy benchmarking of Elixir functions. It also supports comparing sets of benchmarking results.
Comparing benchmarking results is a topic that I have struggled with for years. I run a benchmark several times and get varying results. Then, I make a change to my program and I want to decide if the change causes an improvement in the benchmark score. I rerun the benchmark several times and again get varying results. How do I compare these results? I can compare average score, but is that accurate? How do I tell if the mean of the second run is large enough to be meaningful? How do I know if it is "in the noise?"
Bmark answers this questions using statistical hypothesis testing. Given two sets of benchmark runs, bmark can show:
RunA: RunB:
24274268 6426990
24563751 6416149
24492221 6507946
24516553 6453309
24335224 6491314
24158102 6405073
24357174 6504260
24213098 6449789
24466586 6532929
24289248 6509800
24366622.5 -> 6469755.9 (-73.45%) with p < 0.0005
t = 391.56626146910503, 18 degrees of freedom
This shows that RunA ran in an average of 24366622.5 ms and RunB ran in an average of 6469755.9 ms and that the runtime improved by 73.45% which is statistically meaningful with a confidence level of 99.95%.
## Usage
Add Bmark as a depdency in your mix.exs file:
def deps do
[ {:bmark, "~> 1.0.0"} ]
end
### Writing Benchmarks
To create a benchmark with bmark, create a file ending in `_bmark.ex`. Put the file in a directory called `bmark`. Alltogether, that should look like this
Project Root
+-- bmark
| +-- example_bmark.ex
+-- lib
| +-- your_project_files
+-- mix.exs
In `example_bmark.ex` you should include a module and benchmark function created by using `bmark`
like this:
```elixir
defmodule Example do
use Bmark
bmark :runner do
IO.puts ":runner test is running"
end
bmark :benchmark_with_runs, runs: 5 do
IO.puts "test running 5 times"
end
end
```
The `:runner` benchmark will be run 20 times, the default number of runs. `:benchmark_with_runs` specifies the `:runs` option and will be run only 5 times.
### Running Benchmarks
To run all benchmarks run:
$ mix bmark
This will produce the files
Project Root
+-- results
| +-- example.runner.results
| +-- example.benchmark_with_runs.results
which will contain the run times, in miliseconds, for each run of the benchmark.
### Comparing Benchmark Results
If you have two results files you can compare them by running
$ mix bmark.cmp results/RunA.results results/RunB.results
and bmark will print out the comparison. Here's an example of the comparison with explantions for each section:
RunA: RunB:
24274268 6426990
24563751 6416149
24492221 6507946
24516553 6453309
24335224 6491314
24158102 6405073
24357174 6504260
24213098 6449789
24466586 6532929
24289248 6509800
The section above contains the raw result data presented side-by-side. This is the same data your would get by looking at RunA.results and RunB.results.
24366622.5 -> 6469755.9 (-73.45%) with p < 0.0005
This line shows the change in mean (average) between the two runs. Next, it shows the percentage change and finally the confidence value. You can interpret this as saying there is `1 - p`, or a greater than 99.95% confidence that the change in means is statistically significant. That is, the smaller the value of `p` the more confident you can be in the change in performance.
t = 391.56626146910503, 18 degrees of freedom
The final line shows the `t` value and degrees of freedom. This is the raw statistical data used in Student's t-test to compute the confidence value.
## More Information
Bmark development has been described in detail on my blog [Learning Elixir](http://learningelixir.joekain.com/) you can find all the related posts on the [Bmark page](http://learningelixir.joekain.com/bmark-posts/)
## Contributing
See [Contributing](CONTRIBUTING.md)