R- How to set y-axis individually when using `facet_grid_sc()`

Question

I want to plot the trend of the density (Density.est) of 3 species (Emar,Ecos and Sphyr) over time (2008,2009,2010) for two protection levels of a protected area (RI and RP). As an example of the dataframe I have:

set.seed(123)
Year <- c(2008,2009,2010)
Prot <- c("RI","RP")
species <- c('Emar','Ecos','Sphyr')

df <- data.frame(Year=rep(Year, times = 3, each = 2),
                 Prot=rep(Prot, times = 9, each = 1),
                 Sp=rep(species, times = 1, each = 6),
                 Density.est=rnorm(18,5,2),
                 Density.se=rnorm(18,5,2)-1.5)
df

   Year Prot    Sp Density.est Density.se
1  2008   RI  Emar    3.879049  4.9027118
2  2008   RP  Emar    4.539645  2.5544172
3  2009   RI  Emar    8.117417  1.3643526
4  2009   RP  Emar    5.141017  3.0640502
5  2010   RI  Emar    5.258575  1.4479911
6  2010   RP  Emar    8.430130  2.0422175
7  2008   RI  Ecos    5.921832  2.2499215
8  2008   RP  Ecos    2.469878  0.1266134
9  2009   RI  Ecos    3.626294  5.1755741
10 2009   RP  Ecos    4.108676  3.8067462
11 2010   RI  Ecos    7.448164  1.2237261
12 2010   RP  Ecos    5.719628  6.0076298
13 2008   RI Sphyr    5.801543  4.3529284
14 2008   RP Sphyr    5.221365  2.9098570
15 2009   RI Sphyr    3.888318  5.2902513
16 2009   RP Sphyr    8.573826  5.2562670
17 2010   RI Sphyr    5.995701  5.1431622
18 2010   RP Sphyr    1.066766  4.8772805

Using the next code...

Plot1 <- ggplot(df, aes(x=Year, y=Density.est, group=Prot, color=Prot)) + 
  geom_line() +
  geom_point()+
  geom_errorbar(aes(ymin=Density.est-Density.se, ymax=Density.est+Density.se), width=.2,
                position=position_dodge(0.05)) +
  facet_wrap(~ Sp, scales = "free")

Plot1

I obtain this graph:

However, in my real dataframe, for the species Sphyr, one year I have an extreme value for the density, so I am trying to set y-axis individually for each panel. I found this link solving this problem. I tried to implement what they proposed with the next code:

df$Density.se.min <- df$Density.est-df$Density.se
df$Density.se.max <- df$Density.est+df$Density.se

df.Ecos.min <- min(df$Density.se.min[df$Sp=="Ecos"]) # I calculate the maximum and minimum value of the y-axis for each species.
df.Ecos.max <- max(df$Density.se.max[df$Sp=="Ecos"])

df.Emar.min <- min(df$Density.se.min[df$Sp=="Emar"])
df.Emar.max <- max(df$Density.se.max[df$Sp=="Emar"])

df.Sphyr.min <- min(df$Density.se.min[df$Sp=="Sphyr"])
df.Sphyr.max <- max(df$Density.se.max[df$Sp=="Sphyr"])


scales_y <- list(
  `Ecos` = scale_y_continuous(limits = c(df.Ecos.min, df.Ecos.max), breaks = seq(df.Ecos.min, df.Ecos.max, (df.Ecos.max-df.Ecos.min)/10)),
  `Emar` = scale_y_continuous(limits = c(df.Emar.min, df.Emar.max), breaks = seq(df.Emar.min, df.Emar.max, (df.Emar.max-df.Emar.min)/10)),
  `Sphyr` = scale_y_continuous(limits = c(df.Sphyr.min, df.Sphyr.max), breaks = seq(df.Sphyr.min, df.Sphyr.max, (df.Sphyr.max-df.Sphyr.min)/10))
)

Then, I try to do the plot using this code:

Plot2 <- ggplot(df, aes(x=Year, y=Density.est, group=Prot, color=Prot)) + 
  geom_line() +
  geom_point()+
  geom_errorbar(aes(ymin=Density.est-Density.se, ymax=Density.est+Density.se), width=.2,
                position=position_dodge(0.05)) +
  facet_grid_sc(~ Sp, scales = list(y=scales_y))

Plot2

However, I obtain the error message attempt to select less than one element in get1index.

Does any one know where is the problem?

Answer 1

facet_grid_sc() with "~" does not work. Changing to: + facet_grid_sc(rows= vars(Sp), scales = list(y=scales_y)) seems to work.