...from a geol friend, and goes a long way in explaining why those junior mining market commentators who pretend to be geols but have no qualifications are basically full of shit.
"...here is my chart of the day and an associated diversion from the markets. It shows geochemical data from a project presented in a way to investigate alteration intensity. Each rock type is identified by a symbol. If the world was very simple, each rock type would plot in its own unique location but like people, rock composition is variable right from birth. Added to this original compositional variation are analytical variation and variations from effects of hydrothermal alteration, which is the variation of interest to me. The element concentrations reported by the lab need to be adjusted mathematically based on chemical theory (not statistics) in a manner that helps understand a geological process like alteration. In this one, the scatter in pink and red dots shows how much alteration has modified the rock composition. Since altered rocks are found close to ore ($$$) and unalterated ones are further away, explorers can use this alteration gradient to direct them to the deposit. The pink diamonds started life at about (0.07, 0.7) and the red ones around (0.10, 0.10). Alteration displaces these original compositions up to the right so using this diagram with knowledge of where each point is located in geographic space, I can define the alteration gradient. The trick is knowing how to process the analytical data and arrange them for plotting. In case you are interested, the trick is based on vector algebra...every mineral has a chemical formula (eg. orthoclase, a common mineral in rocks = KAlSi3O8; which shows the molar proportions of elements as one part K, one part Al, 3 parts Si, 8 parts O; or 1:1:3:8). The molar proportions are the coefficents in a matrix and elements are the variables. Since we have many minerals making up the rocks we sample, we have multiple equations of minerals that define a multidimensional geochemical space. These can be solved in a manner that orients the view of space in a way that helps us understand the data. Cool huh?"
I love geologists. DYODD, dude.