denoise
is a script (written in Python 3) which denoises data without using
fitting functions. It implements the algorithm described
here.
All code from this project is licensed under the GPLv3. See the
LICENSE
file for more information.
The following modules are used:
matplotlib
numpy
You can install them with the following command:
pip3 install matplotlib numpy
As an example, running
./denoise -i noisy_data.txt -o denoised_data.txt -u 20.0
will take the data sequence on noisy_data.txt
(one value per line), denoise it
and write the denoised sequence on denoised_data.txt
. Above, -u 20.0
specifies
the denoising coefficient μ: μ = 0.0 means no denoising at all, while larger
values of μ enforce more denoising and therefore generate smoother output
sequences.
You can view a plot of both the original and denoised sequences by adding -p
to the command above.
For testing purposes, sample data with noise is provided on the noisy_data
subdirectory. Try running:
./denoise -i noisy_data/sine.txt -p -u 10.0
and set the denoising coefficient to different values to get a feeling for how the denoising algorithm works.
Diego Assencio / diego@assencio.com