How to make a tidy multiple graph

mhab · March 30, 2021, 8:46pm

Hi All,

Please help me to build this kind of tidy graph (picture as example). I think which type of graph that I should use?
Basically the simple graph theme_bw.

Thank you

williaml · March 30, 2021, 9:16pm

With facet_grid() and facet_wrap() from ggplot2?

vlad_aluas · March 31, 2021, 7:30am

If you can provide a reprex example, we would be able to give you more details. Without more details, you can try using:

your_graph +
facet_wrap( ~ your_variable)

Keep in mind that you will need a specific variable by which to split the data.

If this does not help, please provide a reprex and we can give you more details.

Dobrokhotov1989 · March 31, 2021, 8:44am

What exactly you want to reproduce from this plot? There is a lot of thing going on.

To have a five distinct graphs arranged together, I would make them separately and arrange with cowplot packege (Introduction to cowplot).

Since, your y-axis is different on all individual plots, I don't think facet_ would work in this case.

mhab · March 31, 2021, 9:00am

I would like to produce a y-tick-right-graph, top axis ticks, graph border and the graph close together.
I have tried with theme(axis.ticks.y.right), but not working. Also the legend inside the graph/

mhab · March 31, 2021, 10:14am

p4 <- ggplot() +
geom_line(data = df,aes(x=datetime,y=SWC_1_1_1.x))+
labs(y= "SWC ", x="Datetime")

p4<- p4 + theme(axis.ticks.y.right = element_line(size=5,color="black"),
axis.ticks.length=unit(.2,"cm"),
panel.grid.minor = element_blank(),
panel.background = element_rect(fill = "white", colour = "grey50")+
scale_x_continuous(sec.axis = dup_axis(labels = NULL)) + # NEW
scale_y_continuous(sec.axis = dup_axis(labels = NULL)))
p4

Here is my code.But not working as I expected

williaml · March 31, 2021, 9:13pm

What is in df? Can you provide a reproducible example?

FAQ: How to do a minimal reproducible example ( reprex ) for beginners Guides & FAQs

A minimal reproducible example consists of the following items: A minimal dataset, necessary to reproduce the issue The minimal runnable code necessary to reproduce the issue, which can be run on the given dataset, and including the necessary information on the used packages. Let's quickly go over each one of these with examples: Minimal Dataset (Sample Data) You need to provide a data frame that is small enough to be (reasonably) pasted on a post, but big enough to reproduce your issue. Let's say, as an example, that you are working with the iris data frame head(iris) #> Sepal.Length Sepal.Width Petal.Length Petal.Width Species #> 1 5.1 3.5 1.4 0.…

mhab · March 31, 2021, 9:29pm

Example:

datetime	micro.form	Exp_Flux	SWC_1_1_1.x	Ts_1_1_1.x
31/08/2018 12:48	HP	6.12	0.26	29.25
31/08/2018 12:52	PB	2.89	0.26	29.15
31/08/2018 12:55	FP	3.22	0.261	29.25
31/08/2018 12:59	HP	10.07	0.261	29.2
31/08/2018 13:02	PB	5.87	0.26	29.175
31/08/2018 13:06	FP	2.95	0.259	29.3
31/08/2018 13:09	HP	4.27	0.259	29.225
31/08/2018 13:13	PB	2.27	0.259	29.4
31/08/2018 13:16	FP	3.77	0.261	29.3
31/08/2018 13:20	HP	10.71	0.259	29.4
31/08/2018 13:24	PB	6.15	0.259	29.3
31/08/2018 13:27	FP	3.18	0.258	29.375

Dobrokhotov1989 · April 1, 2021, 3:03am

Something like this?

library(tidyverse)
library(cowplot)

df <- mpg

p4 <- ggplot() +
  geom_line(data = df,
            aes(x=displ,
                y=cty,
                color = as.factor(year)))+
  labs(y= "SWC ", x="Datetime")

p4 <- p4 +
  #Top x axis with no labels and title
  scale_x_continuous(sec.axis = dup_axis(labels = NULL, name = NULL)) +
  #Right y axis with no labels and title 
  scale_y_continuous(sec.axis = dup_axis(labels = NULL, name = NULL)) +
  #Simple theme with wight background and without guides
  theme_classic()+
  theme(
    #Direction of ticks (negative value means it will be inward)
    axis.ticks.length.y.right = unit(-3, "pt"),
    axis.ticks.length.x.top = unit(-3, "pt"),
    #Legend position inside the plot
    legend.position = c(.99, .99),
    legend.justification = c("right", "top"),
    #Line around legend plate    
    legend.background = element_rect(color = "grey50")
    )

# p5 same as p4 but without x-axis labels and title 
p5 <- p4 + 
  labs(x = NULL)+
  theme(axis.text.x.bottom = element_blank())

# arrange them together
plot_grid(p5, p4, ncol = 1)

mhab · April 1, 2021, 8:26pm

Thank you so much.

That really helps.

mhab · April 1, 2021, 9:54pm

Hye..

Unfortunately, another issue came up

Error in as.POSIXct.numeric(value) : 'origin' must be supplied

Dobrokhotov1989 · April 2, 2021, 1:14am

Hmm,

You can try to replace scale_x_continuous() with scale_x_datetime().
If it doesn't work, please run the following code dput(head(df)) where df is data.frame with your data and post the output here (should look like structure(list(....),..., class = ...). It will give me a sample of your data which I can simply copy-paste.

mhab · April 2, 2021, 9:21am

Hye,

I am using (scale_x_datetime) and it's worked..Thank you very much. I wonder what the difference between datetime and continuous?

mhab · April 2, 2021, 9:23am

Dobrokhotov1989 · April 2, 2021, 9:57am

The difference in the class of data. From the name, it implied that scale_*_continuous() is for continuous x or y (i.e. numeric), while scale_*_datetime() is for the data of class datetime (i.e. date and time). There are also scale_*_date() for dates, scale_*_time() for times , scale_*_discrete() for discrete data (i.e. integers, factors, etc.).

Perhaps, the error appeared because ggplot tried to transform data of one class to another, which requires extra argument 'origin'.

mhab · April 2, 2021, 11:10am

Hi,

This is the final output I produce. It was good. Just a little bit I have to add to be the same as sample picture (above).

I want to add a dashed line horizontally across the datetime
And would like to add vertical smooth line.
Here is my code.

p1 <- ggplot(data=df,aes(x=datetime, y=P_Rain_mm))+
geom_bar(stat="identity",width =2.0,color="black")+
labs(y= "Rainfall (mm) ", x="Datetime")
p1
p1<- p1+
#Top x axis with no labels and title
scale_x_datetime(sec.axis = dup_axis(labels = NULL, name =element_blank())) +
#Right y axis with no labels and title
scale_y_continuous(sec.axis = dup_axis(labels = NULL, name=element_blank()))+
theme(
axis.ticks.length.x.top = unit(-5, "pt"),
#Direction of ticks (negative value means it will be inward)
axis.ticks.length.y.right=unit(-5,"pt"),
panel.grid.minor = element_blank(),
panel.background = element_rect(fill = "white", colour = "grey50")
)
p1
p1<- p1 +labs(x = NULL)+
theme(axis.text.x.bottom = element_blank())
p1

p2 <- ggplot(data = df,aes(x=datetime,y=Mg_CO2_ha_yr_exp_flux,group=month(datetime))) +
geom_line(aes(color=micro.form))+
labs(y= "Mg_CO2_ha_yr ", x="Datetime")
p2

p2<- p2+#Top x axis with no labels and title
scale_x_datetime(sec.axis = dup_axis(labels = NULL, name =element_blank())) +
#Right y axis with no labels and title
scale_y_continuous(sec.axis = dup_axis(labels = NULL, name=element_blank()))+
theme(
axis.ticks.length.x.top = unit(-5, "pt"),
#Direction of ticks (negative value means it will be inward)
axis.ticks.length.y.right=unit(-5,"pt"),
#Legend position inside the plot
legend.position = c(.99, .99),
legend.justification = c("right", "top"),
#Line around legend plate
legend.background = element_rect(color = "grey50"),
panel.grid.minor = element_blank(),
panel.background = element_rect(fill = "white", colour = "grey50")
)

p2
p2<- p2 +labs(x = NULL)+
theme(axis.text.x.bottom = element_blank())

p3 <- ggplot() +
geom_line(data = df,aes(x=datetime,y=RH_1_1_1.x, group=month(datetime)))+
labs(y= "Rel_Hum ", x="Datetime")+
theme(legend.position= "none")
p3
p3<- p3 +
#Top x axis with no labels and title
scale_x_datetime(sec.axis = dup_axis(labels = NULL, name =element_blank())) +
#Right y axis with no labels and title
scale_y_continuous(sec.axis = dup_axis(labels = NULL, name=element_blank()))+
theme(
axis.ticks.length.x.top = unit(-5, "pt"),
#Direction of ticks (negative value means it will be inward)
axis.ticks.length.y.right=unit(-5,"pt"),
panel.grid.minor = element_blank(),
panel.background = element_rect(fill = "white", colour = "grey50")
)
p3
p3<- p3 +labs(x = NULL)+
theme(axis.text.x.bottom = element_blank())

p4 <- ggplot() +
geom_line(data = df,aes(x=datetime,y=SWC_1_1_1.x,group=month(datetime)))+
labs(y= "SWC ", x="Datetime")
p4

p4 <- p4 +
#Top x axis with no labels and title
scale_x_datetime(sec.axis = dup_axis(labels = NULL, name =element_blank())) +
#Right y axis with no labels and title
scale_y_continuous(sec.axis = dup_axis(labels = NULL, name=element_blank()))+
theme(
axis.ticks.length.x.top = unit(-5, "pt"),
#Direction of ticks (negative value means it will be inward)
axis.ticks.length.y.right=unit(-5,"pt"),
panel.grid.minor = element_blank(),
panel.background = element_rect(fill = "white", colour = "grey50")
)
p4
p4<- p4 +labs(x = NULL)+
theme(axis.text.x.bottom = element_blank())
p4

p5 <- ggplot() +
geom_line(data = df,aes(x=datetime,y=WTD_1_1_1.x, group=month(datetime)))+
labs(y= "WTD ", x="Datetime")
p5
p5<- p5 +
#Top x axis with no labels and title
scale_x_datetime(sec.axis = dup_axis(labels = NULL, name =element_blank())) +
#Right y axis with no labels and title
scale_y_continuous(sec.axis = dup_axis(labels = NULL, name=element_blank()))+
theme(
axis.ticks.length.x.top = unit(-5, "pt"),
#Direction of ticks (negative value means it will be inward)
axis.ticks.length.y.right=unit(-5,"pt"),
panel.grid.minor = element_blank(),
panel.background = element_rect(fill = "white", colour = "grey50")
)
p5
library(gridExtra)
ggarrange(p1, p2,p3, p4,p5,
ncol = 1, nrow= 5)

Dobrokhotov1989 · April 5, 2021, 2:56am

For horizontal line, you need to use geom_hline() so your code should be something like this:

p1 + 
  geom_hline(yintercept = 1000, linetype = "dashed")

I'm not sure what do you mean by a "vertical" smooth line. Perhaps, you need geom_smooth(method = ...). There are several methods for smoothing, so check the documentation for details.

P.S. I suggest you read a good tutorial about ggplot2. I found this one quite useful: A-ggplot2-tutorial-for-beautiful-plotting-in-r

mhab · April 5, 2021, 8:48pm

Thank you very much!!!!
Really appreciate that!!

system · April 12, 2021, 8:48pm

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.