Main Data Repository for Submission

  

Main Supporting Data

This folder contains all relevant supporting data, images, and documentation for the different components of our submission other than the clustering algorithm component (described above). The most important files and folders are called out in detail below. 

• all_epma_ols – This is a large master CSV containing all measured olivine compositions using the electron microprobe. This spreadsheet also includes some contextual data, like when the analysis was made and what grouping the olivine falls into in our textural classification scheme. 
• Slamet Ni Mass Balance_V2.xlsx – this spreadsheet comprises the different mass balance fractionation models as described in Chapter 5. The first few sheets describe the methods and supporting information supporting the modeling approach. Each individual mass balance model is given its own sheet. 
• SlametLoyangMeltInclusionsv3.xlsx – This spreadsheet comprises all available whole rock, host mineral, and glass compositional and modeling data with regards to each melt inclusion measured in this submission. Such data include pre- ('Database' sheet) and post-PEC ('PEC' sheet) corrected compositions, Raman analyses, SIMS measurements, and entrapment pressures. The first two sheets in this spreadsheet are also designed to be read by Python, hence the column name format and layout. The final sheet presents the data in a human-friendly way

NOTE: Two melt inclusion records have unreasonably high H2O and CO2 contents (marked in red in third sheet of spreadsheet. At this time it is not clear if this is a transcription error on my part or a problem with an Excel formula in an earlier iteration of this database. These faulty MIs have bene removed from our analyses, pending more detailed review. 

• Plots/ - this folder contains all relevant code used to construct most of the figures in Chapter 5. This folder also contains supporting datasets, like filtered spreadsheets showing only glass compositions. The most important code contained in this spreadsheet is “EPMA All Data.ipynb.” 
• Melt-inclusions/ - this folder contains all code, spreadsheets, and .melts scripts used to process and analyze melt inclusion data in Chapter 5. Some important components to call out are 1) melts-pec-corrections/ - this folder contains all rhyoliteMELTS command line files and outputs associated with the “manual” PEC correction protocol described in the methods; 2) MeltInclusionPlotting.ipynb – this python file contains all code used to make melt inclusion plots; 3) tutorials/ contains all relevant code used to perform saturation pressure calculations using MagmaSat.

Machine Learning Code

This folder contains a selection of scripts, input files, and output figures used to perform hierarchical clustering in the manuscript. Below I call out a few key files: 

• NickClusterNotebookV2.ipynb – this file is my adapted version of co-author Felix Boschetty’s 2022 clustering algorithm. All code used to perform and analyze the clusters is available here. The content of the algorithm is the same as described in his paper, but this notebook includes some tweaks like different plotting styles. 
• Slamet_Loyang_NB_v2.xlsx – this spreadsheet contains all relevant core, rim, and profile compositions from olivines in my Slamet/Loyang database. This spreadsheet is a cleaned-up version of the “master” CSV file for olivine compositions discussed above. 
• figs/ folder – contains all finalized and earlier draft versi