ggplot,geom_bar, facet_wrap : How to define specific dates as x values, for each plot

[LauraL]

Hi.
As image illustrates, my barplot are not symmetrical in terms of width of bars, due to different number of observations, and different timespans for the individual years (2015, 2016 and 2018). I want the bars to be of equal size, but defining the width (width=) it becomes very messy.

Besides, for some unknown reason, the dateformat changes in 2018, where i suddenly only have months on the x axis , and not months+day as the previous years.

I've tried everything i could think off, including the functions gap.plot, gap.barplot and axis.break from the plotrix package (v3.7-6), but it dosent work with my ggplot.

library(tidyverse)
library(rmarkdown)
library(readxl)
library(dplyr)
library(tidyr)
library(lubridate)
library(reshape2)
library (ggthemes)
library(ggpubr)
library(plotrix)
library(scales)

#printing dataframe
print(gapl151618_wetCS)

N_ug_m2_h   se_graph      Date      Trt  Year
<dbl>      <dbl>         <date>    <chr> <chr>
0.59172415	1.1529715	2018-06-15	C	2018
0.79847845	0.5764257	2018-07-01	C	2018
-2.02633456	-1.8041389	2018-09-19	C	2018
2.03877800	2.2450879	2018-06-15	S	2018
-1.19586851	-1.9191906	2018-07-01	S	2018
0.00776147	0.5356943	2018-09-19	S	2018
6.99500000	2.2524320	2016-07-07	C	2016
7.97983333	2.6705785	2016-07-18	C	2016
2.98933333	2.3721254	2016-08-01	C	2016
-1.65866667	-1.1795099	2016-08-15	C	2016
5.52157486	1.4964004	2015-08-06	C	2015
5.05983333	3.2670861	2016-07-07	S	2016
7.28733333	1.5233716	2016-07-18	S	2016
3.82466667	3.5809984	2016-08-01	S	2016
0.30750000	1.4447466	2016-08-15	S	2016
-0.09014181	-3.8432439	2015-08-06	S	2015

#creating plot
ggplot(gapl151618_wetCS, aes(x=Date , y=N_ug_m2_h, fill=Trt))+
  geom_hline (yintercept=0, color = "black", linetype = 3 ) +
  facet_wrap(~Year, scales = "free_x") +
  geom_bar(position = position_dodge2 (width = 1), stat = "identity", colour="black", size=.3) +
  geom_errorbar(aes(ymin=N_ug_m2_h, ymax=N_ug_m2_h+se_graph, group = Trt), position = position_dodge2 ()) +
   ylab(expression(~mu~g~N[2]*O-N~~m^{-2}~h^{-1})) +
  xlab("Date") +
  ggtitle(expression(N[2]*O~fluxes~wet~tundra~2015~","~2016~and~2018)) 

Defining width on geom_bar

theme_set(
  theme_bw()+
    theme(legend.position = "right",
          plot.title = element_text (hjust = 0.5, vjust = 1, size = 15),
          axis.text.x = element_text (angle = 60, hjust = 1)))

ggplot(gapl151618_wetCS, aes(x=Date , y=N_ug_m2_h, fill=Trt))+
  geom_hline (yintercept=0, color = "black", linetype = 3 ) +
  facet_wrap(~Year, scales = "free_x") +
  geom_bar(position = position_dodge2 (width = 1), stat = "identity", colour="black", size=.3, width = 1) +
  geom_errorbar(aes(ymin=N_ug_m2_h, ymax=N_ug_m2_h+se_graph, group = Trt), position = position_dodge2 ()) +
    ylab(expression(~mu~g~N[2]*O-N~~m^{-2}~h^{-1})) +
  xlab("Date") +
  ggtitle(expression(N[2]*O~fluxes~wet~tundra~2015~","~2016~and~2018)) 

[andresrcs]

If I correctly understand what you are trying to do, I think you can use factors on the x-axis instead of dates and use facet_grid() with space = "free" argument.

library(ggplot2)

gapl151618_wetCS <- data.frame(stringsAsFactors=FALSE,
   N_ug_m2_h = c(0.59172415, 0.79847845, -2.02633456, 2.038778, -1.19586851,
                 0.00776147, 6.995, 7.97983333, 2.98933333, -1.65866667,
                 5.52157486, 5.05983333, 7.28733333, 3.82466667, 0.3075, -0.09014181),
    se_graph = c(1.1529715, 0.5764257, -1.8041389, 2.2450879, -1.9191906,
                 0.5356943, 2.252432, 2.6705785, 2.3721254, -1.1795099,
                 1.4964004, 3.2670861, 1.5233716, 3.5809984, 1.4447466, -3.8432439),
        Date = as.Date(c("2018-06-15", "2018-07-01", "2018-09-19", "2018-06-15",
                 "2018-07-01", "2018-09-19", "2016-07-07", "2016-07-18",
                 "2016-08-01", "2016-08-15", "2015-08-06", "2016-07-07", "2016-07-18",
                 "2016-08-01", "2016-08-15", "2015-08-06")),
         Trt = c("C", "C", "C", "S", "S", "S", "C", "C", "C", "C", "C", "S",
                 "S", "S", "S", "S"),
        Year = c(2018, 2018, 2018, 2018, 2018, 2018, 2016, 2016, 2016, 2016,
                 2015, 2016, 2016, 2016, 2016, 2015)
)

ggplot(gapl151618_wetCS, aes(x= as.factor(Date), y=N_ug_m2_h, fill=Trt))+
    geom_hline (yintercept=0, color = "black", linetype = 3 ) +
    facet_grid(~Year, scales="free", space = "free") +
    geom_bar(position = position_dodge2 (width = 1), stat = "identity", colour="black", size=.3) +
    geom_errorbar(aes(ymin=N_ug_m2_h, ymax=N_ug_m2_h+se_graph, group = Trt), position = position_dodge2 ()) +
    ylab(expression(~mu~g~N[2]*O-N~~m^{-2}~h^{-1})) +
    xlab("Date") +
    ggtitle(expression(N[2]*O~fluxes~wet~tundra~2015~","~2016~and~2018))

^{Created on 2019-10-15 by the reprex package (v0.3.0.9000)}

Note: Next time, please try to narrow down your code to just the relevant part and provide a proper REPRoducible EXample (reprex)

[Matthias]

Interesting that facet_grid has this option but not facet_wrap, I was looking into the help to find something like this.

[LauraL]

Its a very nice suggestion, many thanks
The problem is, that I still would like to have unequal spacing between bars on x axis (in order to visualize the different time-spans between the sets of observations (C and S), and this feature is lost in this solution. But it is still better than what i made so far:)