Breast-Cancer-Diagnosis
Table of content
The project requires Python 3.6+, and python packages listed in requirements.txt file
- Creating conda virtual environment [in Linux]
conda create -n env
conda activate env
conda config --env --add channels conda-forge
conda config --env --set channel_priority strict
- Cloning the repo and install the dependency packages using
requirements.txt
git clone https://github.com/Caphace-Ethan/Breast-Cancer-Diagnosis
cd Breast-Cancer-Diagnosis
conda install -r requirements.txt or try using pip instead of conda
sudo apt-get install graphviz or conda install python-graphviz
- radius (mean of distances from center to points on the perimeter)
- texture (standard deviation of gray-scale values)
- perimeter
- area
- smoothness (local variation in radius lengths)
- compactness (perimeter^2 / area - 1.0)
- concavity (severity of concave portions of the contour)
- concave points (number of concave portions of the contour)
- symmetry
- fractal dimension ("coastline approximation" - 1)
The mean, standard error (SE) and "worst" or largest (mean of the three largest values) of these features were computed for each image
Package Scripts are found in scripts directory, its content is explained below.
FileHandler:Helper class for reading and writing different file formats.package_config:Script for handling package configurations such as location of files and url(s), etc.data_processor:Script for performing data processing for example computing missing values, fixing outliers, etc.plot:Script for ploting different figures including box-plot, scatter-plot, histogram, bar-plot, etc.log:script for logging package logs