Nuts & Bolts: Minerals | Ternary Mixer | Unit Test | Glyphs
The data file should be a plain text file containing the compositions of interest. The fields can be delimited with either a comma or TAB character. The delimiter will be automatically determined.
The first row is the header row and must be included. It names the components of the compositions. The components can be named as elemental symbols (eg: Si, Ca, Mg...) or as oxide strings (eg: SiO2, CaO, MgO). Capitalisation is essential as some strings are ambiguous without it. The only other valid value for a column header is the atomic number of the element, though we do not expect this to be used very frequently.The header items for the error columns are auto-detected, so while you are welcome to use any string here to identify your error column it is important that the header string does not resolve to an element or oxide. It is OK to use the unicode character for "sigma" (σ) if you wish, this website is entirely UTF-8 compatible.
The first column contains the sample names. These can be as long as you wish, but shorter will appear better in the plots.
Subsequent columns contain the composition in either weight percent or mol fraction. This is automatically detected: if the composition total falls in the range 0 to 2 then mol fraction is assumed, otherwise it will be regarded as weight percent. If you wish to specify the precision of the composition then every second column can hold the error/precision value.
Your data file may contain further columns of related data that are not part of the composition. Any column header that does not resolve to an element symbol or oxide string will be ignored, so for example, if you have a "total" column at the end of your table.
Some experimental sample results expressed as oxides. Error column included. Comma delimeted/CSV file.
The same experimental sample results expressed as oxides, but with the error column excluded. Comma delimeted/CSV file.
Here is the example data file plotted on the ADF ternary. The site can produce much larger files, and plot thousands of compositions to a single chart.