I have treatment and control groups stored separately in two dfs. I am interested in presenting two variables 1) sentiment and by 2) month_year across for the two groups in the same graph. Each row in the df represents a tweet followed by the predicted sentiment and the month_year where it was written. For example, in the control group, the data looks as follows:

tweet sentiment   month_year
xyz   negative.   March_2022
xyz   positive.   March_2022
xyz   neutral.    March_2022
xyz   negative.   April_2022

And similarly, the treatment group df are structured as follows:

tweet sentiment   month_year
xyz   negative.   March_2022
xyz   positive.   March_2022
xyz   positive.    March_2022
xyz   positive.   April_2022

And I am interesting in count the share of negative tweets per month across time and between the two groups.

Here is my attempt at creating the graph for one group. However, I am interested in generating the same indicator below but for both groups at once, so that I can present them in the same graph where I compare the trends for both groups throughout time.

Create a variable counting 1-negative sentiment posts and 2-their share per month

sentiment_monthly <- control_group  %>%
 group_by(month_year) |>
   #group_by(treatment_details) |>
  summarise(sentiment_count = n(),
            negative_count = sum(sentiment_human_coded == "negative"),
            negative_share = negative_count/sentiment_count * 100) 

Here is a data example of the "sentiment_monthly" df:

dput(sentiment_monthly[1:5],)

output:

structure(list(month_year = structure(c(2011.16666666667, 2011.25, 
2011.41666666667, 2011.75, 2011.83333333333, 2011.91666666667, 
2012.08333333333, 2012.16666666667, 2012.25, 2012.33333333333
), class = "yearmon"), sentiment_count = c(272L, 62L, 64L, 434L, 
111L, 59L, 72L, 144L, 43L, 17L), negative_count = c(27L, 23L, 
47L, 317L, 79L, 27L, 25L, 78L, 27L, 3L), negative_share = c(9.92647058823529, 
37.0967741935484, 73.4375, 73.0414746543779, 71.1711711711712, 
45.7627118644068, 34.7222222222222, 54.1666666666667, 62.7906976744186, 
17.6470588235294), year = c(2011, 2011, 2011, 2011, 2011, 2011, 
2012, 2012, 2012, 2012)), row.names = c(NA, -10L), class = c("tbl_df", 
"tbl", "data.frame"))

and then the viz:

Visualizing negative sentiment by month

ggplot(data = sentiment_monthly, aes(x = as.Date(month_year), y = negative_share))  
    geom_bar(stat = "identity", fill = "#FF6666", position=position_dodge())   
    scale_fill_grey()  
    scale_x_date(date_breaks = "1 month", date_labels = "%b %Y")  
    theme(plot.title = element_text(size = 18, face = "bold"))  
  theme_bw() 
   theme(axis.title.x=element_blank(),
                      axis.ticks.x=element_blank())   # remove x-axis label
    theme(plot.title = element_text(size = 5, face = "bold"),
          axis.text.x = element_text(angle = 90, vjust = 0.5))

output: [![enter image description here][1]][1]

Based on the useful advice below, I did this, which works well!

control_graph |> select(month_year,group, negative_share) |> 
  filter(group == "control")

treatment_graph |> select(month_year,group, negative_share) |> 
 filter(group == "treatment")
control_graph |> 
  bind_rows(treatment_graph) |> 
ggplot(aes(x = as.Date(month_year), y = negative_share, fill = group))   
  geom_bar(stat = "identity", position=position_dodge())

CodePudding user response：

You could try to bring the two dataframes in a similar struture, bind_rows and create a grouped bar chart like this:

library(tidyverse)
control <- gapminder::gapminder |> select(year,continent, gdpPercap) |> 
  filter(continent == "Asia")

treatment <-  gapminder::gapminder |> select(year,continent, gdpPercap) |> 
  filter(continent == "Europe")


control |> 
  bind_rows(treatment) |> 
  ggplot(aes(x=year, y = gdpPercap, fill = continent))   
  geom_bar(stat = "identity", position=position_dodge())