Data for: Statistically learning Archean carbonate diagenesis
Here we provide 3 spreadsheets with all the primary data (as received from the from the Laser Ablation ICP-MS facility at University of New Brunswick) LA-ICP-MS (Laser Ablation - Induced Coupled Plasma - Mass Spectrometry) performed on samples of stromatolitic dolostone from the Lower Transvaal Supergroup in Botswana. These files contain the raw data of LA-ICP-MS as received from the laboratory at the University of New Brunswick (Canada). The analyses have been performed on Neoarchean stromatolitic dolostone from the Ramonnedi Formation (lower Transvaal Supergroup) of Botswana. Each file contains 2 spreadsheets, one with results reduced using NCS610 standard and the other one with results reduced using MACS standard. Below is a description of the content of the spreadsheets. Columns A-B: reference material and source file name (internal laboratory code); Columns C-E: date and time of the execution of the analyses Column F: duration of the analyses Column G: name of the sample analyzed Column H: calcium content in Counts Per Second Column I: Internal Standard Error Columns J - BU: trace element concentrations and internal standard errors. LA-ICP-MS analysis has been carried out on thin sections at the Laser Ablation ICP-MS facility at University of New Brunswick (UNB), Fredericton (Canada), using an Agilent 7700x ICP-MS coupled with a Coherent CompexPro 110 (193 nm Excimer laser) and a Resonetics M-50-LR laser ablation system. Carbonate analyses were performed using a 33µm spot size with a repetition rate of 3Hz and an on-sample energy of 5J/cm2, with a 30s ablation and a 30s gas blank between each ablation. Carrier gasses were ultra-pure helium (300 ml/min), ultra-pure nitrogen (2 ml/min), and standard Argon (930 ml/min). The second rotary pump was also used which almost doubles the sensitivities of heavy isotopes. A full suite of elements was monitored during tuning to ensure maximum sensitivity over the range of masses we were analyzing, while keeping doubly charged ions and oxides at a maintainable level (below 0.3% for each). Standards used were NIST610 and MACS-3. Calcium was used as an internal standard for data reduction of carbonate samples. The dwell times for most isotopes were kept at 0.01 sec per isotope, allowing us the lowest possible sweep time for each method. We also provide a zip file with the R codes for the clustering, MANOVA, and discriminant analysis and the table with the primary data in the right format for running the codes.