Search code examples
rggplot2facetaxis-labels

Plotting different x-axis using facet grids

I am plotting spider-diagrams for geochemical analysis.

I managed to plot different series of elements separately (major elements and rare-earth elements) but I would like to plot them together using facet_grid. The problem I get is that the x-axis is common. I would like to have two separate x-axis like I show in the imgur post : https://i.sstatic.net/cmjKW.jpg

I've written commented codes on what I achieved :

library(readxl)
library(tidyr)
library(dplyr)
library(ggplot2)


data <- read_excel("Documents/TFB/xlsx_geochimie/solfa_total_tout_ppm.xlsx", 
                   col_types = c("text", "numeric", "numeric", 
                                 "numeric", "numeric", "numeric", 
                                 "numeric", "numeric", "numeric", 
                                 "numeric", "numeric", "numeric", "numeric", 
                                 "numeric", "numeric", "numeric", 
                                 "numeric", "numeric", "numeric", "numeric", 
                                 "numeric", "numeric", "numeric", "numeric", 
                                 "numeric", "numeric", "numeric", "numeric", "numeric", 
                                 "numeric", "numeric", "numeric", "numeric", "numeric", 
                                 "numeric", "numeric", "numeric", 
                                 "numeric", "numeric", "numeric", "numeric", 
                                 "numeric", "numeric", "numeric", "numeric", 
                                 "numeric", "numeric", "numeric", 
                                 "numeric", "numeric", "numeric", 
                                 "numeric", "numeric", "numeric", "numeric", 
                                 "numeric", "numeric", "numeric", "numeric", 
                                 "numeric", "numeric", "numeric", "numeric", 
                                 "numeric", "numeric", "numeric", "numeric"))

### Vectors containing different geochemical series

vec_maj = c("SiO2","TiO2","Al2O3","FeO","MgO","CaO","Na2O","K2O")
vec_TR = c("La","Ce","Pr","Nd","Sm","Eu","Gd","Tb","Dy","Ho","Er","Tm","Yb","Lu")
vec_tout <- as.character(c(vec_maj,vec_TR))
data.mod <- data[vec_tout]
data.mod$Ech <- data$Ech
### Wide format to long format
data.lf = data %>% select(c(vec_tout,"Ech")) %>%
  pivot_longer(-Ech,names_to="Element",values_to="Pourcentage") %>%
  mutate(Element=factor(Element,levels=unique(vec_tout)))
### Plotting the series separately

data.maj <- subset(data.lf,data.lf$Element %in% vec_maj)
View(data.maj)
data.TR <- subset(data.lf,data.lf$Element %in% vec_TR)

ggplot(data=data.maj,mapping=aes(x=Element,y=Pourcentage,colour=Ech))+
  geom_point()+geom_line(aes(group=Ech))+scale_y_log10()

ggplot(data=data.TR,mapping=aes(x=Element,y=Pourcentage,colour=Ech))+
  geom_point()+geom_line(aes(group=Ech))+scale_y_log10()

# Plotting the series together, x-axis scales does not split :-( 
data.lf$Type <- ifelse(data.lf$Element %in% vec_maj,"Major","REE")
ggplot(data=data.lf,mapping=aes(x=Element,y=Pourcentage,colour=Ech))+
  geom_point()+geom_line(aes(group=Ech))+scale_y_log10()+facet_grid(Type~.,scales="free")

You may download my dataset here : google drive

Solution

  • Instead of facet_grid, can you use facet_wrap?

    ggplot(data.lf, mapping = aes(x = Element, y = Pourcentage, colour = Ech)) +
  geom_point() +
  geom_line(aes(group = Ech)) +
  scale_y_log10() +
  facet_wrap(Type ~ ., ncol = 1, scales = "free")
# Warning: Removed 8 rows containing missing values (geom_point).

    facet wrap