I have some R experience and wanted to learn Python for data science. This book is great for that. Chapters are nicely layer out, starting from preprocessing, feature selection and then machine learning models. It even has a section on Restricted Boltzmann Machines for image analysis.

compared with other data science book in python, this one is thinner but still comprehensive. Not the best if you want to start learning all the tools and methods, but great for reviewing and refreshing what you've learnt from other places

I am a senior engineer with years of experience working primarily in C, C#, perl, and T-SQL. I have basic python, and dusty memories of two years of college math. In the last year, my data set has ballooned at the rate of 1Tb every two months and will soon exceed the handling capacity of my old analytics stack. Blessed by my manager with a shiny new hadoop cluster and time to study, I'm learning new tricks. This book is one of the first I found, and for me it was perfect. It reads like a walk-through from a smart coworker: enough to get me going, the most important moving parts, a few gotchas, where to go for help, some simple working examples... It got me moving on my first project in just a few hours. This is the book I'd have written for myself.

Although I am an experienced Data Scientist who knows well Python's stack for Data Science (scikit-learn, pandas, statsmodels, numpy, scipy, matplotlib, IPython), this book captured my attention and I have read a half of it during the first two days after getting the book. This book is easy to read for novices and experts alike (it does not contain a lot of math and wherever there are formulas they are not difficult to grasp), though some familiarity with Python packages comprising the Data Science stack will greatly facilitate material understanding. The writing style authors chose is excellent as it teaches readers in a very logical and pedagogically appealing way: the way of data pre-processing and analysis occur in projects that data scientists and engineers often encounter when aiming to solve the real-worlds tasks.The books begins with a description of how to install Python and various packages needed to run the code. The purpose of these packages is also explained. Different Python distributions are briefly discussed together with their characteristics, so that a reader can select a distribution particularly suitable to his/her needs. As all code examples in the book are run in IPython Notebook, special attention is paid to a short but comprehensive introduction into IPython itself. Data sets used in the book are described too.After advising on installation of Python and its packages, the book guides readers towards fast and easy data loading from a file, including the case when the entire data set cannot be loaded during one read in the memory and the solution offered is to load it in chunks by using pandas.Furthermore, answers to the following problems are provided: how to deal with erroneous records, how to treat categorical and text data, what are useful data cleansing and transformation operations implemented in pandas, how to use the optimized data structures - numpy arrays - and what operations on them can be done.Once data is loaded and converted to a suitable representation, the book then spends a chapter on the general Data Science pipeline that can be implemented with scikit-learn. The pipeline includes dimensionality reduction via either feature extraction or feature selection, outlier detection, predictive modeling (classification and regression), optimization of model's hyper-parameters, and model's performance evaluation. This material creates the holistic view what typical data analysis is comprised of.The next chapter introduces several popular machine learning algorithms in detail. Among them are linear and logistic regression, Naive Bayes, support vector machines, bagging and boosting ensembles. Special attention is paid to scikit-learn solutions of the 3Vs of big data: namely, volume, velocity and variety. Scalability with volume is solved with incremental learning when at any given moment of time, only a portion (batch) of the entire data fit to the available memory is used to update a model, hence, a model learns incrementally as new batches arrive. To keep up with velocity, scikit-learn offers a number of classification and regression algorithms optimized for speed. Data variety is deal with the help of hashing and sparse matrices. The chapter ends with short examples of doing basic operations of Natural Language Processing with the NLTK package and data clustering.Final two chapters are devoted to social network analysis with the NetworkX package and data visualization with the matplotlib and pandas packages, respectively.Although I have both paper and electronic versions of this book, I would advise first to buy the paper version as numerous code is much easier to understand in this format because one can see the entire snapshot at once.

The book covers fundamentals of Data Science. Code for the book is available from the publisher. I used Anaconda Launcher which nicely converted the notebooks to jupyter and ran them well. My favorite chapter was chapter five Social Network Analysis. I like the table on graph examples, type, node and edges. It is useful for writing code.