CPython Internals: Your Guide to the Python 3 Interpreter

@epythonlab #pythonbooks

How to add an Element to a Set in Python

https://www.pythonforbeginners.com/basics/how-to-add-an-element-to-a-set-in-python

@epythonlab #code #article

K-Centroid Clustering

Summary: Cluster analysis identifies cohesive subgroups of observations within a dataset. It allows us to reduce a large number of observations into a smaller number of clusters.

STEP 1: SELECT APPROPRIATE VARIABLES

The first step is to understand the objectives for segmentation. Then, choose the appropriate variables that provide the information needed for clustering. A sophisticated cluster analysis cannot compensate for the poor choice of attributes.

STEP 2: DATA PREPARATION

Numeric data: Cluster analyses requires numeric data. Many non-numeric variables can be converted to numeric ones. Make sure to remove outliers as clustering algorithms are highly sensitive to outliers.

Variable reduction: This step often requires variable reduction techniques to combine variables that revolve around a particular theme. A common method is Principal Component Analysis (PCA), which reduces a set of related variables into few principal components (PCs) that explain most of the variances in the data. Rule of thumb is to use PCs that account for ~80% variance.

Scaling the data: Standardizing each variable using the z-score ensures that the results are not overly sensitive to variables with higher values.


STEP 3: DETERMINE THE NUMBER OF CLUSTERS

Use the AR and CH indices to determine the optimal method and number of clusters. Use a box and whisker plot. The higher the median and smaller the variation the better. Remember, clustering is an iterative process and may require comparing several models to arrive at a good solution.


STEP 4: CREATE THE CLUSTERING MODEL

Select the variables, standardization process, clustering method, and number of clusters that gave the best solution. Create the cluster model and append the clusters to the dataset.


STEP 5: VISUALIZE AND VALIDATE RESULTS

Visualization helps us determine the meaning and usefulness of the clustering solution. Use summary statistics to understand difference among clusters.

Validate the results: You can use internal validation and/or external validation. Plot the distribution of the validation variable for each cluster using box and whisker plot to visualize the differences.

#keynotes #cluster #kcentroid #dataanalysis @epythonlab

Tableau is the best tool for building interactive Dashboard.
Here is the online free trial for 13 days and you can also download the public
https://dub01.online.tableau.com/#/site/pa/home

Check If a List has Duplicate Elements

https://www.pythonforbeginners.com/basics/check-if-a-list-has-duplicate-elements

@epythonlab #list #article

Scraping is one way of obtaining data from web page. Here you can learn Web scraping with some python libraries.

Mastering Web Scraping in Python: Scaling to Distributed Crawling

https://www.zenrows.com/blog/mastering-web-scraping-in-python-scaling-to-distributed-crawling

@epythonlab #article #code #webscraping

The process of from raw data to story.

BTW, I wish a happy new year for #Ethiopian.

Don't miss it out
Link: http://www.udacity.com/scholarships/nokia-bit-scholarship

It may help you to get understand how clustering algorithm and some other forecasting tools implemented in this project.

Essential Python for the Physicist - 2020

@epythonlab #pythonbooks #physics

Morioh is a social networking platform for programming and that connects programmers abroad.
https://morioh.com/p/eb78f8b88dcc
@epythonlab

Completely Free SMS Gateway

Do you know how to create fake data?
https://youtu.be/0HyIwcZBV3U

STEM scholarship for Women's #udacity #scholarship
Link: https://bit.ly/3AS6Rlm

For more info, join @epythonlab