CODEWITHSUNDEEP
rggplot2lapplyposixct
saiful
saiful

Reputation: 13

Recursive ggplotting of POSIXct data with CORRECT corresponding plot name

need help: My original df has 50 vars: 49 POSIXct class and 1 categorical outcome var. I want to plot (ggplot) the 49 vars to explore its time distribution as a scatterplot. I want to put the plots into a grid of plots. The trouble is getting the corresponding plot to var name correct. My solution below seems longwinded. 

```{r}
library(grid)
library(ggplot2)
library(gridExtra)
str(a[c(1:4,50)])
```

'data.frame': 48 obs. of 4 variables:

$ admd : POSIXct, format: "2017-09-23 12:00:00" "2017-09-23 10:31:00" ...

$ feverd: POSIXct, format: "2017-09-23 12:00:00" "2017-09-23 12:00:00" ...

$ defd : POSIXct, format: "2017-09-23 16:00:00" "2017-09-23 12:13:00" ...

$ ns1d : POSIXct, format: NA "2017-09-23 10:13:00" ...

$ outcome: Factor w/ 2 levels "Death","Discharge": 2 1 2 2 2 2 2 2 2 2 ...

I want to plot these: intended results/plots: each plot has its corresponding variable name

What I did isn't polished at all:

#plot_fun2
```{r}
plot_fun<- function(df){
  ggplot(df, aes_string(x=df[[1]], y=rnorm(1:48))) +
    geom_point(aes(col=a$outcome), alpha=0.5, cex=3)+
    theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
  xlab("Time") + ylab("RandomVisualValue")+
  scale_colour_manual(name="Outcome",values=c("#FF420E","#89DA59")) +
  ggtitle(names(df)[[1]])
}
```

```{r}
sample_list<-list(
  data.frame(a[1]), data.frame(a[2]), data.frame(a[3]), data.frame(a[4])) #CAN WE SIMPLIFY THIS? my original data frame has 49 vars!

timedistb<-lapply(sample_list, plot_fun)
do.call(grid.arrange, c(timedistb, ncol=2, nrow=2))
```

Snippet of data: https://github.com/dcicantab5/recover-study/blob/master/c.csv

Upvotes: 0

Views: 251

Answers (1)

adibender
adibender

Reputation: 7568

Don't really understand what the random number y-axis is about, but you may want to try some combination of gather and facet_wrap:

  • gather: gathers the date time variables in one column datevar that stores column names and one column dateval that stores the respective dates (from each row).
  • facet_wrap: creates facets for each column name in datevar, where each facets only contains data from the respective date time variables (previous column names of date time variables)

Example code:

library(dplyr)
library(tidyr)
library(ggplot2)
df <- read.csv("https://raw.githubusercontent.com/dcicantab5/recover-study/master/c.csv")
df <- df %>% 
    gather(datevar, dateval, -X, -outcome) %>% 
    mutate(rnum = rnorm(1))
ggplot(df, aes(x=dateval, y=rnum)) +
    geom_point(aes(col=outcome), alpha=0.5, cex=3)+
    theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
  xlab("Time") + ylab("RandomVisualValue")+
  scale_colour_manual(name="Outcome",values=c("#FF420E","#89DA59")) + 
  facet_wrap(~datevar)

Upvotes: 1

Related Questions