Skip to main content

A simple module for outlier detection thanks to Modified Thompson Tau Test

Project description

Installation

mt3 requires Python 3.8+

To install the package run :

pip install mt3

If you are planing to use it with numpy and/or pandas, add optionnal dependencies :

pip install mt3[pandas, numpy] # or pip install mt3[numpy] for numpy only

mt3 will then be capable to deal with numpy.ndarray and pd.Series.

By default mt3 is provided with a table of Student T critical values. Available confidence levels are [0.9, 0.95, 0.975, 0.99, 0.995, 0.999]. To be able to use any confidence level, add scipy optionnal dependency :

pip install mt3[scipy]

Usage

mt3 main function is modified_thompson_tau_test :

from mt3 import modified_thompson_tau_test

sample = [-4, 3, -5, -2, 0, 1, 1000]

# You can use it with a simple list :

modified_thompson_tau_test(sample, 0.99)
# [False, False, False, False, False, False, True]


# But you can also use it with a numpy.ndarray or a pandas.Series
import numpy as np
import pandas as pd

modified_thompson_tau_test(np.array(sample), 0.99)
# [False False False False False False True] (numpy array)

modified_thompson_tau_test(pd.Series(sample), 0.99)
# [False False False False False False True] (numpy array)

# If you have nan values in your array or Series, you can choose to treat
# them as outliers
sample_with_nan = np.array([-4, np.nan, 3, -5, -2, 0, 1, 1000])

modified_thompson_tau_test(sample_with_nan, 0.99, nan_is_outlier=True)
# [False True False False False False False True] (numpy array)

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

mt3-0.2.1.tar.gz (6.9 kB view details)

Uploaded Source

Built Distribution

mt3-0.2.1-py3-none-any.whl (7.2 kB view details)

Uploaded Python 3

File details

Details for the file mt3-0.2.1.tar.gz.

File metadata

  • Download URL: mt3-0.2.1.tar.gz
  • Upload date:
  • Size: 6.9 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: poetry/1.1.13 CPython/3.8.10 Windows/10

File hashes

Hashes for mt3-0.2.1.tar.gz
Algorithm Hash digest
SHA256 8ae38a496e415630b08c0688b03cabc9ea8a13ce31a77c880bc8a36d4b670e71
MD5 48e0da37ad07eab329c66b00822d1fe8
BLAKE2b-256 0dc4e22ce89cfcb227e26cd9f1a2984d867867c9f73118f386a536084873439f

See more details on using hashes here.

Provenance

File details

Details for the file mt3-0.2.1-py3-none-any.whl.

File metadata

  • Download URL: mt3-0.2.1-py3-none-any.whl
  • Upload date:
  • Size: 7.2 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: poetry/1.1.13 CPython/3.8.10 Windows/10

File hashes

Hashes for mt3-0.2.1-py3-none-any.whl
Algorithm Hash digest
SHA256 71cedb414d29f175ecd48d9032feb9e568e686d16a5736b90e08e52fc3e1fbc1
MD5 6eb79a9cff16faf56151407bc1deb32d
BLAKE2b-256 579be80b1bc18b2b11cdef601c5d40082a643e08d96c2ff1a60395ca80b4137d

See more details on using hashes here.

Provenance

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page