Introduce science to SQL optimization.
Profiles the performance of SQL queries by running the N times and measuring the time it takes to execute with minimizing any external reasons such as compilation time, queue time e.t.c.
How many times SQL query will be run is determined by confidence level and margin of error.
Compares results of 2 different queries using Student's t-test. Returns the result of the test (p-value) and the conclusion.
Table of currently supported databases:
| Data Base | Current status |
|---|---|
| Redshift Serverless | Supported ✅ |
| Redshift Dedicated | In development ⏳ |
| BigQuery | Planned |
| PostgreSQL | Planned |
| MySQL | Planned |
CLI is in development. For now, you can use the package in your code.
To install package run
pip install bluescope
To use the package in your code:
from bluescope import REDSHIFT_SERVERLESS
from bluescope import get_profiler
from bluescope.statsutils import find_significance
profiler_rs_cls = get_profiler(REDSHIFT_SERVERLESS)
profiler_rs = profiler_cls(host=*host*, port=*port*,
db=*db*, user=*user*,
password=*password*, agree=True)
pr_1 = profiler_rs.profile(*query_1*)
pr_2 = profiler_rs.profile(*query_2*)
p = find_significance(pr_1['mean'], pr_2['mean'],
pr_1['std'], pr_2['std'],
pr_1['sample_size'], pr_2['sample_size'])Any ideas, comments, code improvements are very welcome.
To contribute to the project, please follow the steps:
- Fork the repository
- Create a new branch
- Make your changes
- Push your changes to your branch
- Create a pull request
- Wait for the review
⚖️ GPL-3.0
Maintainer: Mirzabekian Arkadii 📧 Email