Skip to contents

ASTR schema: Naming conventions

Source: vignettes/VG.ASTR.Schema.0.0.2.Rmd
VG.ASTR.Schema.0.0.2.Rmd

The ASTR schema defines a set of naming and data conventions to create data objects that can be easily subset with (almost) no loss of information. Such datasets allow frictionless handling of them and their subsets between software and humans. Creating such datasets requires reduction of implicit information as much as possible (e.g. units of a value being defined in a separate column). Moreover, data objects must be readable for machines while preserving human readability and allow for direct import of spreadsheet-like formats. The conventions defined below are further influenced by considerations related to the analytical data and their quality, which are explained in this vignette.

The schema is platform and programming language agnostic. You can learn about the implementation of the schema in the ASTR package in the dedicated vignette.

Data preparation

  • All variable names are case sensitive.
  • Decimals can only be read when indicated by a decimal point (0.5), as opposed to decimal commas (0,5).
  • Apostrophes and any other special characters should be avoided in the sample/variable names.

Analytical data columns naming

All columns that contain element and oxide compositions, isotopic values, or ratios derived therefrom as well as analytical uncertainties (“errors”) should be named using the following conventions:

  • Only Latin characters can be supported in the column names, i.e. δ and ε will not be identified.
  • The names of oxides and trace elements are self-explanatory (e.g. SiO2 or Si). Total iron, if given, should be expressed as: FeOtot or Fe2O3tot. Loss on ignition, where known, should be expressed as LOI. The value “balance” provided by e.g. many pXRF instruments should be expressed as Bal or Balance.
  • Units for values should follow element or oxide, as _<unit>. Units supported include all SI units, as well as ppm, ppb, ppt, ppq, %, wt%, at%, ‰, counts, and cps, noted as such.
  • Where known, specify wt% and at% instead of using %, and make the mixture type of concentrations explicit, e.g. write ppm(m/m) instead of ppm. For ‘per mille’, use the symbol ‰.
  • For isotopic ratios, use simple forms such as 206Pb/204Pb, 87Sr/86Sr, 147Sm/144Nd, or d18O. For isotope ratios expressed in δ-notation and similar expressions, always provide the mass of the isotope in the numerator (e.g. write e143Nd instead of eNd).
  • If your dataset includes columns with additional information such as geolocation, or values derived from analytial data such as Pb isotope age model parameters, ensure the names do not contain a dash (‘-’) or underscore (‘_’) prior to import and specify these columns as context in the read_ASTR() function.
  • Analytical precision should be indicated in the column name as: _err2SD, _errSD, _errSE, _err2SE without indicating the unit. For absolute analytical uncertainties, the unit will be inferred from the corresponding composition column. Relative analytical uncertainties are indicated by adding a % sign, e.g. _err2SD%. We strongly recommend that the uncertainty is included in the data table, and properly noted following the conventions described.
Components Accepted formats
Oxides and elements SiO2 ; Si
Total iron FeOtot ; Fe2O3tot
Loss on ignition LOI
Isotopic ratios 206Pb/204Pb ; 87Sr/86Sr ; 147Sm/144Nd ; d18O
Units for value _wt% ; _at%
Analytical precision _err2SD ; _errSD ; _errSD% ; _errSD% ; _errSE ; _err2SE