In maxgav13/GMisc: Miscellaneous Functions for Geology (and its areas)

library(kableExtra)
library(GMisc)
library(plotly)
library(ggplot2)

knitr::opts_chunk$set(
  echo = TRUE,
  message = FALSE,
  warning = FALSE,
  error = FALSE,
  fig.align = "center",
  out.width = "80%"
)

Diagrams

This document shows how to display and use different classification diagrams that are used frequently in geosciences (for more info on each one you can click the function name to go to the help page) these are (diagram - function):

• QAP - ternary_qap()

• QAP for gabbros - ternary_qap_g()

• QAP for mafic rocks - ternary_qap_m()

• QAP for ultramafic rocks - ternary_qap_um()

• QtFL for provenance - ternary_qtfl()

• QmFLt for provenance - ternary_qmflt()

• AFM - ternary_afm()

• Pyroclastic rocks - ternary_pyroclastic()

• Folk for sandstone classification - ternary_folk()

• Shepard for soil classification - ternary_shepard()

• Total Alkali Silica (TAS) - TAS()

• Piper diagram - piper_diagram()

Basic use

Static

A simple static diagram without data can be obtained by just calling the respective function. The result is a ggplot object. For the ternary diagramas is based on the ggtern package.

#| fig.cap: Static Folk ternary diagram for sandstones

ternary_folk()

Dynamic

To obtain a dynamic diagram the output argument must be set to 'plotly', resulting in a plotly object.

#| fig.cap: Dynamic QAP ternary diagram for mafic rocks

ternary_qap_m(output = 'plotly')

#| fig.cap: Dynamic TAS diagram

TAS(output = 'plotly')

Add your data!!

To add your own data to the diagram a new table must be created (imported), where the names of the columns should preferably match the names of the axes for the corresponding diagram, or map them accordingly. To do this check the Names section where the respective names are given for each diagram, and to what axis they have to be mapped to.

In the case of the ternary diagrams, the sample (row) totals (adding the three components) are not required to add-up to 100%, this is done internally.

Adding some data is demostrated here by ploting two points on the QtFL and TAS diagrams, first by ploting the data in a static diagram (ggplot), and then ploting the same data in the dynamic diagram (plotly).

In the data for the ternary diagram (d) the first sample (point) adds-up to a 100 but the second one does not, showing the point made above.

d = data.frame(sample=c('a','b'),
               qt=c(40,30),
               f=c(40,20),
               l=c(20,10))

d2 = data.frame(silica=c(58,70), 
                alkali=c(8,4))

#| echo: false
d %>% 
  kable(caption = 'Samples for QtFL diagram') %>% 
  kable_classic()

#| echo: false
d2 %>% 
  kable(caption = 'Samples for TAS diagram') %>% 
  kable_classic()

Static

For the ggplot object, data can be added using the function geom_point, where the only argument that is required is data, the rest can be omitted or modified.

#| fig.cap: Static QtFL ternary diagram for provenance, with user's data.

ternary_qtfl() + 
  geom_point(data = d,
             color = 'coral',
             size = 3,
             shape = 3,
             alpha = .7)

#| fig.cap: Static TAS diagram, with user's data.

TAS() + 
  geom_point(aes(x=silica, y=alkali),
             data = d2,
             color = 'coral',
             size = 3,
             shape = 3,
             alpha = .7)

Dynamic

For the plotly object it is a bit different for the bivariate and ternary diagrams.

For the ternary diagrams, data can be added using the function add_trace, where the arguments data, a, b, c, type = "scatterternary", and mode = "markers" are required, the rest can be changed. The vertices a, b, and c, correspond to the top, bottom left, and bottom right of the respective ternary diagram. The name is what gets displayed in the legend, and the marker defines the appearance.

#| fig.cap: Dynamic QtFL ternary diagram for provenance, with user's data.

ternary_qtfl(output = 'plotly') %>% 
  add_trace(a = ~qt, b = ~f, c = ~l,
            data = d,
            name = 'My data',
            type = "scatterternary",
            mode = "markers",
            marker = list(size = 8,
                          color = 'green',
                          symbol = 4,
                          opacity = .7),
            hovertemplate = paste0('Qt: %{a}<br>',
                                   'F: %{b}<br>',
                                   'L: %{c}'))

For the bivariate diagrams, data can be added using the function add_markers, where the arguments data, x, and y are required, the rest can be changed. The name is what gets displayed in the legend, and the marker defines the appearance. The layout(showlegend = TRUE) forces the legend to show up.

#| fig.cap: Dynamic TAS diagram, with user's data.

TAS('plotly') %>%
  add_markers(x = ~silica, y = ~alkali,
              data = d2,
              name = "My data",
              marker = list(size=8,color='orange',symbol=3,opacity=.9)) %>%
  layout(showlegend = TRUE)

Piper diagram

The Piper diagram is a special case since the data needs pre-processing in order to be plotted on the piper_diagram() function. The following example shows how to do this by using the piper_data_prep() function. This diagram is only available as static (ggplot) object.

d3 = data.frame(Group = c('A','A','B','B'),
                Ca = c(120,150,110,52.6),
                Mg = c(78,160,110,28),
                Na = c(210,590,340,51.6),
                K = c(4.2,2,3.6,2.3),
                HCO3 = c(181,181,189,151),
                CO3 = 0,
                Cl = c(220,744,476,72.2),
                SO4 = c(560,1020,584,126))

piper_data = piper_data_prep(d3)
piper_data %>% 
  kable(caption = 'Processed sample data for Piper diagram') %>% 
  kable_classic()

Once the data is transformed it can be plotted using the x and y values as the coordinates. As with the other diagrams, further customization of the resulting object is posible.

#| fig.cap: Piper diagram, with user's data.

piper_diagram() +
   geom_point(aes(x,y,col=Group,shape=Group),
              size=3,
              data = piper_data) +
   scale_color_brewer('Group',palette = 'Dark2') +
   scale_shape_manual('Group',values = c(3,21))

Names {#tb-names}

These are the suggested names that your data should have for you to be able to plot your data in the corresponding diagram, focusing on the ternary diagrams. For the TAS diagram the names of the axes are x and y. Other names can be used but they must be mapped to the corresponding axis.

#| echo: false

diagrams.tern = data.frame(
    diagram = c('afm','folk','pyroclastic','qap_g','qap_m','qap_um','qap','qmflt','qtfl','shepard'),
    x = c('a','f','lapilli','cpx','ol','opx','a','f','f','sand'),
    y = c('f','q','bb','p','p','ol','q','qm','qt','clay'),
    z = c('m','r','ash','opx','px','cpx','p','lt','l','silt'),
    a = c('~f','~q','~bb','~p','~p','~ol','~q','~qm','~qt','~clay'),
    b = c('~a','~f','~lapilli','~cpx','~ol','~opx','~a','~f','~f','~sand'),
    c = c('~m','~r','~ash','~opx','px','~cpx','~p','~lt','~l','~silt')
)

diagrams.tern %>% 
  kable(caption = 'Mapping of the data to the respective axis for both outputs, for the ternary diagrams.') %>% 
  add_header_above(c('','ggplot'=3,'plotly'=3)) %>% 
  kable_classic()

Español (Spanish)

Todos los diagramas se pueden generar en español, para ello es necesario usar el argumento language y ponerlo igual a 'es', por defecto se despliegan en inglés ('en'). [A spanish version is available for all the diagrams, to display this version the language argument must be set to 'es', by default is set to english ('en').]

#| fig.cap: Diagrama ternario de Shepard para la clasificación de suelos, en español

ternary_shepard(language = 'es')

#| fig.cap: Diagrama TAS, en español

TAS(language = 'es')

Conclusion

This is very basic and quick introduction on how to use the functions to plot diagrams used in geosciences. Further customization of the diagramas can be achived by manipulating the ggplot or plotly objects, specially when adding your own data.

maxgav13/GMisc documentation built on June 12, 2022, 3:48 a.m.