All notable changes to the software will be documented in this Changelog.
The format is based on Keep a Changelog and adheres to Semantic Versioning.
- Added support for parallel GPU training using
fastai.distributed
- Named the library the
Toolkit for Ionospheric Disturbance Detection (TIDD)
to better reflect the purposes of the software and deep learning model. - The supported Python versions to Python 3.7 and 3.8 only.
- Refactored the code to pull existing functionality contained in Jupyter notebooks to
the
tidd
core.
- Renamed
data.normalize
todata.rescale
to better reflect the purpose and functionality of the function. - Adds a notebook for Exploratory Data Analysis called
notebooks/Exploratory Data Analysis of 2012 Hawaii Dataset.ipynb
and removes other not needed notebooks.
- The function
data.normalize
which scales data on a specified scale (such as [0, 1]). This function can be utilized in many analysis and modeling workflows, where rescaling data is often needed for applying specific approaches or algorithms. - A notebook that covers some Exploratory Data Analysis, located at
notebooks/Exploration of Geographic Patterns.ipynb
.
- A
src/data.py
, which contains functions for reading and manipulating data that is available for analysis and experimentation. - A unit testing suite that uses pytest
and pytest-cov to
ensure the software is working as intended, along with a
test_data.py
file that tests the functionality ofsrc/data.py
. - Added an Apache 2.0 License.
- The project
readme.md
to include information about additional datasets available for the 2015 Chilean Earthquake. This data is new data that is in addition to previous available data which covered a 2012 Hawaiian event.