Measuring the Confidence of the k-Nearest Neighbors Algorithm developed in Python for Visualizing the 3D Stratigraphic Architecture of the Llobregat River Delta in NE Spain
Confidence maps for the KNN predictions in https://www.mdpi.com/2077-1312/10/7/986 / https://github.com/dcabezas98/knn-stratigraphic-visualization.
The code can be found in the repository, it can be downloaded as ZIP by clicking in the geen Code button. The only necessary file is the notebook confidence.ipynb.
The data can be found in the data folder. Only two files are necessary: deltacontourn.csv, that contains the points of the contour of the Delta; and hsd new basements.xls, that contains the data from the wells.
You can run the notebook in Jupyter-Notebook or Visual Studio Code, you also need a Python kernel installed in your computer. We recommend installing Anaconda and launching Jupyter by typing jupyter-notebook in the Anaconda Prompt.
To successfully run the notebook, you need to locate it in the same folder as the data directory. In order to do this, you may just extract the ZIP file with the whole repository. Then, launch Jupyter Notebook and select the notebook confidence.ipynb. To run a cell, you can just click in the run button (next to the cell number) or click on it and press Ctrl+Enter. You're now ready to go!
Manuel Bullejos, David Cabezas, Manuel Martín-Martín and Francisco Javier Alcalá