Кластерный анализ (k-means) на python

Кластерный анализ (k-means) на python

k-means, Python. Python , C , Java Python . , , k-means . , , . , : k means, . get_data , file open, numpy.array map . , . Python 3.x. : import numpy as np def get_data: source row.strip.split; for row in openexport.csv names row0 for row in source1: data mapfloat, row1: for row in source1: return names, normdata def normdata: matrix np.arraylistdata, f len_val lenmatrix1, : for i in rangelen_val: local_min matrix:, i.min if local_min 0.0: matrix:, i - local_min local_max matrix:, i.max if local_max 0.0: matrix:, i local_max return matrix.tolist if __name__ __main__: get_data

- Pandas SciKit Learn . : import pandas as pd from sklearn.cluster import KMeans from sklearn.preprocessing import MinMaxScaler, StandardScaler read data drop last empty column, caused by an extra last colon in the header data pd.read_csvfilename, sep;.dropnaaxis1, howall normalize data X MinMaxScaler.fit_transformdata.drop, 1 clustering n_clusters 3 km KMeansn_clustersn_clusters fit predict clusters datacluster km.fit_predictX results - we should have 3 clusters: 0,1,2 printdata clusters centroids printkm.cluster_centers_ : In 47: data Out47: 1 2 3 4 5 6 7 8 9 10 11 0 91.76.155.87 2 2 1 5 3 4 1 1 2 2 1 1 188.123.252.12 2 1 1 6 6 1 2 1 2 3 4 2 46.242.13.93 3 3 1 4 1 2 5 3 1 5 1 3 94.19.223.144 2 2 1 5 6 4 1 1 1 2 4 4 46.39.46.15 2 2 2 5 4 3 3 4 3 2 4 12 13 14 15 16 17 18 cluster 0 3 5 4 5 6 6 4 2 1 5 3 3 4 5 5 5 0 2 5 4 5 5 5 6 4 1 3 5 2 4 6 6 6 5 2 4 4 4 3 3 4 5 3 0 In 48: printkm.cluster_centers_ 0. 0.25 0.5 0.75 0.8 0.33333333 0.375 0.5 0.75 0.16666667 1. 0.75 0.5 0. 0.16666667 0.25 0. 0.5 1. 1. 0. 0. 0. 0.33333333 1. 0.66666667 0. 1. 0. 1. 0.66666667 1. 0.66666667 0.5 1. 0.5 0. 0.5 0. 0.5 0.7 1. 0. 0. 0.25 0. 0.5 0.5 0.5 0.5 0.83333333 1. 1. 0.75

Комментарии

Отправить комментарий

Популярные сообщения из этого блога

Skipping acquire of configured file 'contrib/binary-i386/Packages' as repository … doesn't support architecture 'i386'

Skipping acquire of configured file 'contrib/binary-i386/Packages' as repository … doesn't support architecture 'i386' I installed Ubuntu 18.04 and then while installing Virtualbox i get the i386 error message after I did: sudo apt-get update Get:1 http:nl.archive.ubuntu.comubuntu bionic InRelease 242 kB Hit:2 http:security.ubuntu.comubuntu bionic-security InRelease Hit:3 http:ppa.launchpad.netdanielrichter2007grub-customizerubuntu bionic InRelease Hit:4 http:nl.archive.ubuntu.comubuntu bionic-updates InRelease Hit:5 http:nl.archive.ubuntu.comubuntu bionic-backports InRelease Get:6 http:download.virtualbox.orgvirtualboxdebian bionic InRelease 4429 B Get:7 http:download.virtualbox.orgvirtualboxdebian bioniccontrib amd64 Packages 1426 B Fetched 248 kB in 1s 209 kBs Reading package lists... Done N: Skipping acquire of configured file contribbinary-i386Packages as repository http:download.virtualbox.orgvirtualboxdebian bionic InRelease doesnt support architecture i386 ...
Далее...

Connection string for MariaDB using ODBC

Connection string for MariaDB using ODBC What is the connection string for using MariaDB with ODBC Im guessing its something like DRIVERMariaDB ODBC 5.1 Driver; SERVERLOCALHOST; PORT3306; DATABASEDatabaseName; UIDroot; PASSWORDpassword;OPTION3; Connection strings seems to show strings for MySQL but not MariaDB. MariaDB seems to have a lot of connection strings for Access, SQL Lite, excel, etc ... What have I missed Although this is late answer, but MariaDB have that documented here SQLWCHAR ConnStr LDriverMariaDB ODBC 2.0 Driver;Serverlocalhost;UIDodbc_user;PWDodbc_pw;DBodbc_test;Port3306;
Далее...

Celery like system based on django channels

Celery like system based on django channels According to the docs of django channels: http:channels.readthedocs.ioenlatestconcepts.htmlhighlightcelerynext-steps One thing channels do not do, however, is guarantee delivery. If you need certainty that tasks will complete, use a system designed for this with retries and persistence e.g. Celery, or alternatively make a management command that checks for completion and re-submits a message to the channel if nothing is completed rolling your own retry logic, essentially. I worked with celery some years ago. Yes, it is great, it is big, it is too much. I would like to avoid it. I am searching a simple tasks queue with guarantee delivery and a re-submission on failure. An other alternative would be python-rq. But this has other disadvantages like using pickle data format instead of json. A friend gave me an advice: https:github.comfurious-lukedjango-cq Looks good. From the docs An attempt to implement a distributed task queue ...
Далее...