Self Organizing Map or Hierarchical Cluster Analysis


I am trying to run cluster analysis on my data to see the relative abundance of the different elements. For example, I know that Chlorite is made out of Mg, Fe, Si, Al. I also know that I have nine minerals in my rocks.

So I want to input my elemental data with depth and I want the software to group my data into nine groups that represent the different minerals. Below are my minerals and next to each one are the elements that make up that rock. The numbers in each element represent the number of cations in that mineral. So basically in Chlorite there should be about equal amounts of Mg and Fe (37.5% each), 18.75% of Si and 6.25% of Al. The data should orginize it self in clusters where each one represents a mineral. I understand that there will be complixities but I am sure that there will be ones that will be clear.

Chlorite Mg (6) Fe (6) Si (3) Al (1)

Pyrite Fe (1)  S (2)

Kaolinite Al (2) Si (2)

Dolomite Mg (1) Ca (1)

Quartz Si (1)

Calcite Ca (1)

K-spar Si (3) K (1) Al (1)

Plagioclase Si (2) Al (2) Na (1) Ca (1)

Illite Al (6) Si (4) Mg (2) Fe (2) K (1)

Smectite Si (4) Al (2) Mg (2) Na (0.3) Ca (0.3)

Is that doable? Is there a way where I can create a library of these and after running the analysis it checks with it and it suggests a name?

My excel sheet is organized where I have Columns of depth and the elements (Al, Si, Fe, Mg, Na, K Ca, S).


1 Comment

Would you have a sample dataset with dummy data? I have a bit of a hard time to understand what the detailed data looks like.

akooy - Feb 19, 2019 - 2:34am
+ Add a Comment