Megan Squire - Clean Data [2015, ePub, ENG]

Примеры страниц

Оглавление

1: Why Do You Need Clean Data?
A fresh perspective
The data science process
Communicating about data cleaning
Our data cleaning environment
An introductory example
Summary
2: Fundamentals – Formats, Types, and Encodings
File formats
Archiving and compression
Data types, nulls, and encodings
Summary
3: Workhorses of Clean Data – Spreadsheets and Text Editors
Spreadsheet data cleaning
Text editor data cleaning
An example project
Summary
4: Speaking the Lingua Franca – Data Conversions
Quick tool-based conversions
Converting with PHP
Converting with Python
The example project
Summary
5: Collecting and Cleaning Data from the Web
Understanding the HTML page structure
Method one – Python and regular expressions
Method two – Python and BeautifulSoup
Method three – Chrome Scraper
Example project – Extracting data from e-mail and web forums
Summary
6: Cleaning Data in PDF Files
Why is cleaning PDF files difficult?
Try simple solutions first – copying
Another technique to try – pdfMiner
Third choice – Tabula
When all else fails – the fourth technique
Summary
7: RDBMS Cleaning Techniques
Getting ready
Step one – download and examine Sentiment140
Step two – clean for database import
Step three – import the data into MySQL in a single table
Step four – clean the & character
Step five – clean other mystery characters
Step seven – separate user mentions, hashtags, and URLs
Step eight – cleaning for lookup tables
Summary
8: Best Practices for Sharing Your Clean Data
Preparing a clean data package
Documenting your data
Setting terms and licenses for your data
Publicizing your data
Summary
9: Stack Overflow Project
Step one – posing a question about Stack Overflow
Step two – collecting and storing the Stack Overflow data
Step three – cleaning the data
Step four – analyzing the data
Step five – visualizing the data
Step six – problem resolution
Moving from test tables to full tables
Summary
10: Twitter Project
Step one – posing a question about an archive of tweets
Step two – collecting the data
Step three – data cleaning
Step four – simple data analysis
Step five – visualizing the data
Step six – problem resolution
Moving this process into full (non-test) tables
Summary