Preface

The world of computing has seen an incredible revolution in the past 30 years. Not so long ago, high-performance computations required expensive hardware; proprietary software costing hundreds, if not thousands, of dollars; knowledge of computer languages such as FORTRAN, C, or C++; and familiarity with specialized libraries. Even after obtaining the proper hardware and software, just setting up a working environment for advanced scientific computing and data handling was a serious challenge. Many engineers and scientists were forced to become operating systems wizards just to be able to maintain the toolset required by their daily computational work.

Scientists, engineers, and programmers were quick to address this issue. Hardware costs decreased as performance went up, and there was a great push to develop scripting languages that allowed integration of disparate libraries through multiple platforms. It was in this environment that Python was being developed in the late 1980s, under the leadership of Guido Van Rossum. From the beginning, Python was designed to be a cutting-edge, high-level computer language with a simple enough structure that its basics could be quickly learned even by programmers who are not experts.

One of Python's attractive features for rapid development was its interactive shell, through which programmers could experiment with concepts interactively before including them in scripts. However, the original Python shell had a limited set of features and better interactivity was necessary. Starting from 2001, Fernando Perez started developing IPython, an improved interactive Python shell designed specifically for scientific computing.

Since then, IPython has grown to be a full-fledged computational environment built on top of Python. One of most exciting developments is the IPython notebook, a web-based interface for computing with Python. In this book, the reader is guided to a thorough understanding of the notebook's capabilities in easy steps. In the course of learning about the notebook interface, the reader will learn the essential features of several tools, such as NumPy for efficient array-based computations, matplotlib for professional-grade graphics, pandas for data handling and analysis, and SciPy for scientific computation. The presentation is made fun and lively by the introduction of applied examples related to each of the topics. Last but not least, we introduce advanced methods for using GPU-based parallelized computations.

We live in exciting computational times. The combination of inexpensive but powerful hardware and advanced libraries easily available through the IPython notebook provides unprecedented power. We expect that our readers will be as motivated as we are to explore this brave new computational world.