Measure deviant noise
Project description
measure-noise
Measure how our data deviates from normal distribution
Branch | Status | Coverage |
---|---|---|
master | ||
dev |
Install
pip install measure-noise
Usage
The deviance()
method will return a (description, score)
pair describing how the samples deviate from a normal distribution, and by how much. This is intended to screen samples for use in the t-test, and other statistics, that assume a normal distribution.
SKEWED
- samples are heavily to one side of the meanOUTLIERS
- there are more outliers than would be expected from normal distributionMODAL
- few samples are near the mean (probably bimodal)OK
- no egregious deviation from normalN/A
- not enough data to make a conclusion (akaOK
)
Example
from measure_noise import deviance
>>> desc, score = deviance([1,2,3,4,5,6,7,8])
>>> desc
'OK'
Development
git clone https://github.com/mozilla/measure-noise.git
cd measure-noise
pip install -r requirements.txt
pip install or tests/requirements.txt
python -m unittest discover tests
Windows
You must download the scipy
and numpy
binary packages.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.