Python scientific ecosystem

There are a lot of very good Python packages for sciences. The fundamental packages are in particular:

• Numpy: numerical computing with powerful numerical arrays objects, and routines to manipulate them.

• Scipy: high-level numerical routines. Optimization, regression, interpolation, etc.

• Matplotlib: 2D-3D visualization, “publication-ready” plots.

With IPython and Spyder, Python plus these fundamental scientific packages constitutes a very good alternative to Matlab, that is technically very similar (using the libraries Blas and Lapack). Matlab has a Just-In-Time (JIT) compiler so that Matlab code is generally faster than Python. However, we will see that Numpy is already quite efficient for standard operations and other Python tools (for example PyPy, Cython, Numba, Pythran, …) can be used to optimize the code to reach the performance of optimized Matlab code.

The advantages of Python over Matlab are its open-source nature, its high polyvalency (and nicer syntax) and its huge open-source ecosystem:

• sympy for symbolic computing,

• pandas, statsmodels, seaborn for statistics,

• h5py, h5netcdf for hdf5 and netcdf files,

• mpi4py for MPI communications,

• opencv, scikit-image, pyopencl for image processing,

• pycuda and numba for GPU computing,

• scikit-learn, keras, tensorflow, pytorch, mxnet for machine learning,

• bokeh for display data efficiently,

• mayavi for 3D visualization,

• qtpy, kivy for GUI frameworks

• …

SciPy or NumPy ?

Scipy also provides a submodule for linear algebra scipy.linalg. It provides an extension of numpy.linalg.

For more info, see the related FAQ entry.