pygram11 0.3.0

Fast histogramming in python built on pybind11 and OpenMP.

pygram11

Simple and fast histogramming in python via pybind11 and (optionally) OpenMP.

pygram11 provides fast functions for calculating histograms (and their sums-of-weights squared). The API is very simple, check out the documentation (you'll also find some benchmarks there) and my blog post.

Installing

pygram11 requires only NumPy and pybind11 (and therefore a C++ compiler with C++11 support).

From conda-forge

conda install pygram11 -c conda-forge

Builds on conda-forge will have OpenMP available.

From PyPI

Note: When using PyPI (or source), pybind11 must be installed explicitly before pygram11 (because setup.py uses pybind11 to determine include directories; not an issue if using the conda-forge build).

$ pip install pybind11 ## or `conda install pybind11`
$ pip install pygram11

From Source

$ git clone https://github.com/drdavis/pygram11.git
$ pip install pybind11
$ cd pygram11
$ pip install .

In Action

A fixed bin width histogram of weighted data in one dimension, accelerated with OpenMP:

>>> x = np.random.randn(10000)
>>> w = np.random.uniform(0.8, 1.2, 10000)
>>> h, sw2 = pygram11.histogram(x, bins=40, range=(-4, 4), weights=w, omp=True)
>>> stat_err = np.sqrt(sw2)

A variable bin width histogram in two dimensions:

>>> x = np.random.randn(10000)
>>> y = np.random.randn(10000)
>>> xbins = [-2.0, -1.0, -0.5, 1.5, 2.0]
>>> ybins = [-3.0, -1.5, -0.1, 0.8, 2.0]
>>> h = pygram11.histogram2d(x, y, bins=[xbins, ybins])

Other Libraries

• There is an effort to develop an object oriented histogramming library for python called boost-histogram
• Simple and fast histogramming in Python using the NumPy C API: fast-histogram

---

If there is something you'd like to see in pygram11, please open an issue or pull request.