/CDP

Primary LanguagePython

CDP

Composition Data Process with GUI

This program contains four Composition data transformation algorithms：

  • Improved-alr
  • Clr
  • ilr
  • Stability based on Geboy's approach

It can also generate hierarchical clustering graphs based on these transformation.(The methods for calculating the distance between the newly formed cluster: average, distance metric: correlation)

Usage

1. Run with Python environment (recommend)

  1. install Python-3.7，click here to download python. Note: Select Add Python 3.x to PATH
  2. Run pip install -r requirements.txt
  3. Run python Composition_data_process_gui.py or Double-click the startup script

默认显示语言为英文，若使用简体中文需在启动时加上参数1：python Composition_data_process_gui.py 1。 如果使用脚本启动，用记事本之类的打开修改一下即可。

2. Executable File

I will provide an executable File, for windows only. You can download from releases

methods and metrics

Click: scipy.cluster.hierarchy.linkage

Note: average=UPGMA, weighted=WPGMA, median=WPGMC

Input File

Raw Composition Data Excel File(s)

Raw Data Format

  1. The file is xlsx or xls format.
  2. The first row should be the element name.
  3. The first column should be the sample number.
  4. The second column should be Ash.
  5. The column immediately following Ash is oxide elements, followed by non-oxide elements.

The paper's raw data is in the RawCoalGeochemistryData folder.

Output Files

  • Data for Clr transformation
  • Data for Improved Alr transformation
  • Data for Ilr transformation
  • Stability using Geboy's approach
  • Correlation between element based on improved alr
  • Correlation between element based on Clr approach
  • Correlation between element based on ilr approach
  • Hierarchical cluster graph between element based on Improved-alr, Clr, Ilr, Stability. Support eps, png, svg, pdf, ps