Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/sigma-py/accupy
:dart: Accurate sums and dot products for Python.
https://github.com/sigma-py/accupy
accuracy engineering mathematics numerical-methods numpy pypi python
Last synced: 30 days ago
JSON representation
:dart: Accurate sums and dot products for Python.
- Host: GitHub
- URL: https://github.com/sigma-py/accupy
- Owner: sigma-py
- License: gpl-3.0
- Created: 2018-02-16T19:41:13.000Z (over 6 years ago)
- Default Branch: main
- Last Pushed: 2021-09-06T15:54:02.000Z (about 3 years ago)
- Last Synced: 2024-09-27T04:22:24.100Z (about 1 month ago)
- Topics: accuracy, engineering, mathematics, numerical-methods, numpy, pypi, python
- Language: Python
- Homepage:
- Size: 1.07 MB
- Stars: 103
- Watchers: 3
- Forks: 15
- Open Issues: 2
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
- awesome-scientific-computing - accupy - Accurate sums and dot products for Python. (Other libraries and tools / Mesh tools)
README
Accurate sums and (dot) products for Python.
[](https://pypi.org/project/accupy)
[](https://pypi.org/pypi/accupy/)
[](https://doi.org/10.5281/zenodo.1185173)
[](https://github.com/nschloe/accupy)
[](https://pypistats.org/packages/accupy)
[](https://discord.gg/hnTJ5MRX2Y)
[](https://github.com/nschloe/accupy/actions?query=workflow%3Aci)
[](https://codecov.io/gh/nschloe/accupy)
[](https://github.com/psf/black)
### Sums
Summing up values in a list can get tricky if the values are floating point
numbers; digit cancellation can occur and the result may come out wrong. A
classical example is the sum
```
1.0e16 + 1.0 - 1.0e16
```
The actual result is `1.0`, but in double precision, this will result in `0.0`.
While in this example the failure is quite obvious, it can get a lot more
tricky than that. accupy provides
```python
p, exact, cond = accupy.generate_ill_conditioned_sum(100, 1.0e20)
```
which, given a length and a target condition number, will produce an array of
floating point numbers that is hard to sum up.
Given one or two vectors, accupy can compute the condition of the sum or dot product via
```python
accupy.cond(x)
accupy.cond(x, y)
```
accupy has the following methods for summation:
- `accupy.kahan_sum(p)`: [Kahan
summation](https://en.wikipedia.org/wiki/Kahan_summation_algorithm)
- `accupy.fsum(p)`: A vectorization wrapper around
[math.fsum](https://docs.python.org/3/library/math.html#math.fsum) (which
uses Shewchuck's algorithm [[1]](#references) (see also
[here](https://code.activestate.com/recipes/393090/))).
- `accupy.ksum(p, K=2)`: Summation in K-fold precision (from [[2]](#references))
All summation methods sum the first dimension of a multidimensional NumPy array.
Let's compare them.
#### Accuracy comparison (sum)
As expected, the naive
[sum](https://docs.python.org/3/library/functions.html#sum) performs very badly
with ill-conditioned sums; likewise for
[`numpy.sum`](https://docs.scipy.org/doc/numpy/reference/generated/numpy.sum.html)
which uses pairwise summation. Kahan summation not significantly better; [this,
too, is
expected](https://en.wikipedia.org/wiki/Kahan_summation_algorithm#Accuracy).
Computing the sum with 2-fold accuracy in `accupy.ksum` gives the correct
result if the condition is at most in the range of machine precision; further
increasing `K` helps with worse conditions.
Shewchuck's algorithm in `math.fsum` always gives the correct result to full
floating point precision.
#### Runtime comparison (sum)
We compare more and more sums of fixed size (above) and larger and larger sums,
but a fixed number of them (below). In both cases, the least accurate method is
the fastest (`numpy.sum`), and the most accurate the slowest (`accupy.fsum`).
### Dot products
accupy has the following methods for dot products:
- `accupy.fdot(p)`: A transformation of the dot product of length _n_ into a
sum of length _2n_, computed with
[math.fsum](https://docs.python.org/3/library/math.html#math.fsum)
- `accupy.kdot(p, K=2)`: Dot product in K-fold precision (from
[[2]](#references))
Let's compare them.
#### Accuracy comparison (dot)
accupy can construct ill-conditioned dot products with
```python
x, y, exact, cond = accupy.generate_ill_conditioned_dot_product(100, 1.0e20)
```
With this, the accuracy of the different methods is compared.
As for sums, `numpy.dot` is the least accurate, followed by instanced of `kdot`.
`fdot` is provably accurate up into the last digit
#### Runtime comparison (dot)
NumPy's `numpy.dot` is _much_ faster than all alternatives provided by accupy.
This is because the bookkeeping of truncation errors takes more steps, but
mostly because of NumPy's highly optimized dot implementation.
### References
1. [Richard Shewchuk, _Adaptive Precision Floating-Point Arithmetic and Fast
Robust Geometric Predicates_, J. Discrete Comput. Geom. (1997), 18(305),
305–363](https://doi.org/10.1007/PL00009321)
2. [Takeshi Ogita, Siegfried M. Rump, and Shin'ichi Oishi, _Accurate Sum and Dot
Product_, SIAM J. Sci. Comput. (2006), 26(6), 1955–1988 (34
pages)](https://doi.org/10.1137/030601818)
### Dependencies
accupy needs the C++ [Eigen
library](http://eigen.tuxfamily.org/index.php?title=Main_Page), provided in
Debian/Ubuntu by
[`libeigen3-dev`](https://packages.ubuntu.com/search?keywords=libeigen3-dev).
### Installation
accupy is [available from the Python Package Index](https://pypi.org/project/accupy/), so with
```
pip install accupy
```
you can install.
### Testing
To run the tests, just check out this repository and type
```
MPLBACKEND=Agg pytest
```