Python was able to become the best language for data analysis and mining areas, has its unique advantage. Because he has a lot in this area related to the library can be used, but useful, such as Numpy, SciPy, Matploglib, Pandas, ScikitLearn, Keras, Gensim etc.
1) Numpy, it gives a true Python provides an array of features including multi-dimensional arrays, and functions for fast data processing, Numpy or more advanced extensions library dependencies, such as follow-up Scipy, Matplotlib, Pandas, etc., are the same ;
2) Scipy, he let Python became half MATLAB, Scipy provides true matrix type, and a large number of objects and functions based on matrix operations, he includes features include optimization, linear algebra, integration, interpolation, and you, special functions, fast Fourier transform, signal processing and image processing, solving ordinary differential equations and other scientific and engineering calculations are commonly used; Scipy dependent on Numpy;
3) Matplotlib, Python for it, Matplotlib is drawing the most famous libraries, mostly two-dimensional graphics, of course, can also support a number of short-answer measurements drawing;
4) Pandas, he is the most powerful data analysis and exploration tools under Python, not one. Advanced data structures and sophisticated tools contained him, so that data processing in Python is very fast and simple, Pandas built on NumPy, he makes to Numpy-centric applications are easy to use, comes from the name Pandas panel data ( Panel data) and Python data analysis (data analysis), and he was initially developed as a financial data analysis tool developed by AQR Capital Management company in April 2008, out and open source and the end of 2009;
He's very powerful, supports SQL-like data CRUD, and with a wealth of data processing functions, time series analysis support, support flexible data processing indeed. Pandas actually very complex, enough to write a book alone, if there is interest in him Pandas can look at one of the main authors WesMcKinney wrote, "the use of Python for data analysis," a book.
5) StatModels, Pandas focus on reading, processing and exploration data, and StatsModels is more emphasis on statistical modeling and analysis of data, so that he had the taste of R Python language. StatModels support interaction with the Pandas data, therefore, he combined with Pandas, became a powerful combination of data mining under Python;
6) Scikit-Learn, which is a machine and learn about the library, he is a powerful Python and learning kits, and he provided the perfect learning toolbox, including: data pre-processing, classification, regression, poly class, forecasting and model analysis. He relies on NumPy, SciPy, Matplotlib and so on;
7) Keras, he is used to build neural networks, he is not a simple neural network library, but based on Theano powerful deep learning library that not only can build a common neural network, you can also build a variety of depths learning models, such as from the encoder, recurrent neural network, recurrent neural network, a convolutional neural network. Because it is based Theano, the speed is quite fast.
8) Theano, he is also a Python library, he was led by the depth learning experts YoshuaBengio laboratories developed to define, optimize, and efficiently solve the problem of multidimensional array simulation to estimate the corresponding mathematical expressions. He has implemented efficiently decomposed symbols, highly optimized speed, stability and other characteristics, the most important thing is to achieve a further accelerated GPU processing speed is ten times the data-intensive of the CPU;
9) Gensim, topic modelling of humans, he mainly used the language processing tasks, such as text similarity calculation, LDA, Word2Vec other tasks in these areas often need more background knowledge is often the case: research in this area readers do not need me to say anything without the reader research in this area, here do not know.