Eis-toolkit

EIS Toolkit beta release 3

EIS Toolkit's beta release includes most of the basic tools required to conduct mineral prospectivity mapping with Python. Note that this is a beta release and more tools will be added in the upcoming months, and existing tools will be optimized and refined.


✨ New features

- MLP classifier and regressor (https://github.com/GispoCoding/eis_toolkit/pull/252)
- Surface derivatives (https://github.com/GispoCoding/eis_toolkit/pull/255)
- Filter and smoothing functions (https://github.com/GispoCoding/eis_toolkit/pull/289)
- CLI implementations for many tools


🐞 Fixes

- Fix/improve normality test (https://github.com/GispoCoding/eis_toolkit/pull/267)
- Fix/improve correlation and covariance matrix functions (https://github.com/GispoCoding/eis_toolkit/pull/268)
- Fix/improve PCA (https://github.com/GispoCoding/eis_toolkit/pull/269)


🛠️ Other improvements

- Change raster grid check to use only raster metadata/profile instead of raster dataset (https://github.com/GispoCoding/eis_toolkit/pull/281)
- Added PySAL and ESDA Python packages to repository (https://github.com/GispoCoding/eis_toolkit/pull/288)
- Minor clean-up PR (https://github.com/GispoCoding/eis_toolkit/pull/295)
- Small notebook fixes (https://github.com/GispoCoding/eis_toolkit/pull/296)
- Documentation additions for filter functions (https://github.com/GispoCoding/eis_toolkit/pull/303)
- Documentation for derivative functions (https://github.com/GispoCoding/eis_toolkit/pull/301)


Roadmap
The first non-beta release of EIS Toolkit is planned for April 2024. All the planned tools should be implemented by then and requested new additions from Toolkit testers considered as much as possible. For planned tools, refer to the wiki page of this repository.

The first release of EIS QGIS Plugin, which serves a graphical user-interface for EIS Toolkit, should also happen in April 2024.

EIS Toolkit beta release 2

EIS Toolkit's beta release includes most of the basic tools required to conduct mineral prospectivity mapping with Python. Note that this is a beta release and more tools will be added in the upcoming months, and existing tools will be optimized and refined.

:sparkles: New features

- Unique grid conditions (86)
- One-hot encoding (224)
- Extract shared lines (241)
- Coda transforms (244)
- Calculate geometry (248)
- Logistic regression (210)
- Random forest classifier and regressor (210)
- Gradient boosting classifier and regressor (210)
- Various modeling functions (split data, get cross-validator, load/save model, evaluate model, predict with model) (210)


🚀 Performance improvements

- Optimize IDW (226)
- Optimize distance computation (230)


🐞 Fixes

- Fix invalid CLI imports (distance computation, descriptive statistics) (223)
- Fix several issues with example notebooks (225)
- Fixed issue related to type checking in check_minmax_position function (242)
- Fixed two minor issues with Chi-square test (249)
- Corrections to raster exporting in CBA (250)

🛠️ Other improvements
- Added nodata handling (dropping) in Descriptive statistics (245)
- Miscellaneous ci and documentation additions (256)
- Update fuzzy overlay docs (258)

Roadmap

The first non-beta release of EIS Toolkit is planned for April 2024. All the planned tools should be implemented by then and requested new additions from Toolkit testers considered as much as possible. For planned tools, refer to the wiki page of this repository.

The first release of EIS QGIS Plugin, which serves a graphical user-interface for EIS Toolkit, should also happen in April 2024.

EIS Toolkit beta release

EIS Toolkit's beta release includes most of the basic tools required to conduct mineral prospectivity mapping with Python. Note that this is a beta release and more tools will be added in the upcoming months, and existing tools will be optimized and refined.

Tools included in the release

Conversions

- CSV to dataframe
- Raster to dataframe

Exploratory analysis

- Basic plotting functions from Seaborn (barplot, boxplot, edfcplot, heatmap, histplot, kdeplot, lineplot, pairplot, regplot, scatterplot)
- DBSCAN
- Descriptive statistics (min, max, mean, median, quantiles, standard deviation, relative standard deviation, skew)
- Feature importance evaluation
- K-means clustering
- Plot parallel coordinates
- PCA
- Statistical tests (independence, normality, correlation, covariance)

Prediction

- Fuzzy overlay
- Weights of evidence

Raster processing

- Clip raster
- Create constant raster
- Extract values from raster
- Reproject raster
- Resample raster
- Snap rasters
- Unify rasters
- Extract window from raster

Training data tools

- Balance data (SMOTETomek)

Transformations

- Binarize transform
- Clip transform
- Z-score normalization
- Min-max scaling
- Log transforms (log2, log10, ln)
- Sigmoid transform
- Winsorize

Validation

- Calculate AUC (area under curve)
- Calculate base metrics (true positive rate, proportion of area and false positive rate values)
- Plot correlation matrix
- Plot prediction area curves
- Plot rate curve

Vector processing

- CBA (cell-based association)
- Distance computation
- IDW interpolation
- Kriging interpolation
- Rasterize vector
- Reproject vector
- Vector density


Command-line interface

- Most of the included tools are also available to use from command-line. To get started with this, use the command `eis --help` after installation. Note that the command-line interface has been primarily designed to be an entry-point for other programs and not for direct use


Roadmap

The first non-beta release of EIS Toolkit is planned for April 2024. All the planned tools should be implemented by then and requested new additions from Toolkit testers considered as much as possible. For planned tools, refer to the wiki page of this repository.

The first release of EIS QGIS Plugin, which serves a graphical user-interface for EIS Toolkit, should also happen in April 2024.

v_test
Release created for initial testing purposes.