Python数据分析（影印版）

Python是一种多范式的编程语言，既适合面向对象的应用开发，也适合函数式设计模式。Python已然成为数据科学家们在数据分析、可视化和机器学习方面的**语言，它可以带来高效率和高生产力。
《Python数据分析（影印版 英文版）》将教会初学者如何发掘Python的*大潜力用于数据分析，包括从数据获取、清洗、操作、可视化以及存储到复分析和建模等一切相关主题。它聚焦于一系列开源Python模块，比如NumPy、SciPy、matplotlib、pandas、I Python、Cython、scikit-learn以及NLTK等。在后面的章节里，《Python数据分析（影印版 英文版）》涵盖了数据可视化、信号处理与时间序列分析、数据库、可预测分析及机器学习等主题。

Preface
Chapter 1： Getting Started with Python Libraries
Software used in this book
Installing software and setup
On Windows
On Linux
On Mac OS X
Building NumPy SciPy， matplotlib， and IPython from source
Installing with setuptools
NumPy arrays
A simple application
Using IPython as a shell
Reading manual pages
IPython notebooks
Where to find help and references
Summary

Chapter 2： NumPy Arrays
The NumPy array object
The advantages of NumPy arrays
Creating a multidimensional array
Selecting NumPy array elements
NumPy numerical types
Data type objects
Character codes
The dtype constructors
The dtype attributes
One-dimensional slicing and indexing
Manipulating array shapes
Stacking arrays
Splitting NumPy arrays
NumPy array attributes
Converting arrays
Creating array views and copies
Fancy indexing
Indexing with a list of locations
Indexing NumPy arrays with Booleans
Broadcasting NumPy arrays
Summary

Chapter 3： Statistics and Linear Algebra
NumPy and SciPy modules
Basic descriptive statistics with NumPy
Linear algebra with NumPy
Inverting matrices with NumPy，
Solving linear systems with NumPy
Finding eigenvalues and eigenvectors with-NumPy
NumPy random numbers
Gambling with the binomial distribution
Sampling the normal distribution
Performing a normality test with SciPy
Creating a NumPy-masked array
Disregarding negative and extreme values
Summary

Chapter 4： pandas Primer
Installing and exploring pandas
pandas DataFrames
pandas Series
Querying data in pandas
Statistics with pandas DataFrames
Data aggregation with pandas DataFrames
Concatenating and appending DataFrames
Joining DataFrames
Handling missing values
Dealing with dates
Pivot tables
Remote data access
Summary

Chapter 5： Retrieving， Processing， and Storing Data
Writing CSV files withNumPy and pandas
Comparing the NumPy ．npy binary format and pickling
pandas DataFrames
Storing data with PyTables
Reading and writing pandas DataFrames to HDF5 stores
Reading and writing to Excel with pandas
Using REST web services and JSON
Reading and writing JSON with pandas
Parsing RSS and Atom feeds
Parsing HTML with Beautiful Soup
Summary

Chapter 6： Data Visualization
matplotlib subpackages
Basic matplotlib plots
Logarithmic plots
Scatter plots
Legends and annotations
Three-dimensional plots
Plotting in pandas
Lag plots
Autocorrelation plots
Plot．ly
Summary

Chapter 7： Signal Processing and Time Series
statsmodels subpackages
Moving averages
Window functions
Defining cointegration
Autocorrelation
Autoregressive models
ARMA models
Generating periodic signals
Fourier analysis
Spectral analysis
Filtering
Summary

Chapter 8： Working with Databases
Lightweight access with sqlite3
Accessing databases from pandas
SQLAIchemy
Installing and setting up SQLAIchemy
Populating a database with SQLAIchemy
Querying the database with SQLAIchemy
Pony ORM
Dataset - databases for lazy people
PyMongo and MongoDB
Storing data in Redis
Apache Cassandra
Summary

Chapter 9： Analyzing Textual Data and Social Media
Installing NLTK
Filtering out stopwords， names， and numbers
The bag-of-words model
Analyzing word frequencies
Naive Bayes classification
Sentiment analysis
Creating word clouds
Social network analysis
Summary
Chapter 10： Predictive Analytics and Machine Learning
A tour of scikit-learn
Preprocessing
Classification with logistic regression
Classification with support vector machines
Regression with ElasticNetCV
Support vector regression
Clustering with affinity propagation
Mean Shift
Genetic algorithms
Neural networks
Decision trees
Summary

Chapter 11： Environments Outside the Python Ecosystem and Cloud Computing
Exchanging information with MATLAB/Octave
Installing rpy2
Interfacing with R
Sending NumPy arrays to Java
Integrating SWIG and NumPy
Integrating Boost and Python
Using Fortran code through f2py
Setting up Google App Engine
Running programs on PythonAnywhere
Working with Wakari
Summary

Chapter 12： Performance Tuning， Profiling， and Concurrency
Profiling the code
Installing Cython
Calling C code
Creating a process pool with multiprocessing
Speeding up embarrassingly parallel for loops with Joblib
Comparing Bottleneck to NumPy functions
Performing MapReduce with Jug
Installing MPI for Python
IPython Parallel
Summary

Appendix A： Key Concepts
Appendix B： Useful Functions
matplotlib
NumPy
pandas
Scikit-learn
SciPy
scipy．fftpack
scipy．signal
scipy．stats
Appendix C： Online Resources
Index

《Python数据分析（影印版）》：
Installing and exploring pandas
The minimal dependency set requirements for pandas is given as follows：
NumPy： This is the fundament alnumerical array package that we installed and covered extensively in the preceding chapters
python—dateuh I：Thisis a date—handlinglibrary
pytz： This handles time zone definitions
This list is the bare minimum； a longer list of optional dependencies can be locatedat http：／／pandas.pydata.org／pandas—docs／stable／install.html.We caninstall pandas via PyPI with pip or easy_install， using a binary installer， with theaid of our operating system package manager， or from the source by checking outthe code.The binary installers can be downloaded from http：／／pandas.pydata.org／getpandas.html.
The command to install pandas with pip is as follows：
pip install pandas
You may have to prepend the preceding command with sudo if your user accountdoesn't have sufficient rights.For most， if not all， Linux distributions， the pandaspackage name is python—pandas.Please refer to the manual pages of your packagemanager for the correct command to install.These commands should be the same asthe ones summarized in Chapter 1， Getting Started with Python Libraries.To install fromthe source， we need to execute the following commands from the command line：
$ git clone git：／／github.com／pydata／pandas.git
$ cd pandas
$ python setup.py install
This procedure requires the correct setup of the compiler and other dependencies；therefore， it is recommended only if you really need the most up—to—date versionof pandas.Once we have installed pandas， we can explore it further by addingpandas—related lines to our documentation—scanning script pkg_check.
……