# Cumulative averages in barchart

Hi,
I have a data where I have to track attendance of students school-wise. The tracking happens week-wise. So for each school, on an average, what are the attendance figures is what I need. But the graph that I made shows the wrong numbers. I calculated the cumulative average. In the Y-axis, I gave cumulative average. How can I solve this so that I can see a particular school X's attendance in a particular week?

``````library(tidyverse)

data<-tibble::tribble(
~school,    ~week, ~student_attendance,
"GHPS UNKAL", "week 1",                 69L,
"GHPS VIDYANAGAR", "week 1",                 89L,
"GHPS NARENDRA", "week 1",                 67L,
"GHPS HEGGERI", "week 1",                 88L,
"GHPS UNKAL", "week 2",                 77L,
"GHPS VIDYANAGAR", "week 2",                 80L,
"GHPS NARENDRA", "week 2",                 69L,
"GHPS HEGGERI", "week 2",                 69L,
"GHPS UNKAL", "week 3",                 85L,
"GHPS VIDYANAGAR", "week 3",                 67L,
"GHPS NARENDRA", "week 3",                 79L,
"GHPS HEGGERI", "week 3",                 68L,
"GHPS UNKAL", "week 4",                 66L,
"GHPS VIDYANAGAR", "week 4",                 88L,
"GHPS NARENDRA", "week 4",                 76L,
"GHPS HEGGERI", "week 4",                 87L
)

graph1<-data %>%
mutate(week=as.character(week)) %>%
group_by(school,week) %>%
summarise(avg_attend=mean(student_attendance)) %>%
mutate(cum_avg=cummean(avg_attend))
#> `summarise()` has grouped output by 'school'. You can override using the `.groups` argument.
ggplot(graph1,aes(week,cum_avg))+
geom_bar(width = 0.5,stat = "identity")+
theme_minimal()+
labs(title = "School-wise attendance",
x="week",
y="attendance")

Created on 2022-10-18 by the reprex package (v2.0.1)
``````

This happens because week 1, week 2 etc are repeated for every school:

You could solve using `facet_wrap()`:

``````graph1 %>%
ggplot(aes(week, cum_avg))+
geom_bar(width = 0.5,stat = "identity")+
theme_minimal()+
labs(title = "School-wise attendance",
x="week",
y="attendance") +
facet_wrap(~school)

``````

Or `fill = school` and `position = "dodge"`

``````graph1 %>%
ggplot(aes(week, cum_avg, fill = school))+
geom_bar(width = 0.5,stat = "identity", position = "dodge")+
theme_minimal()+
labs(title = "School-wise attendance",
x="week",
y="attendance")
``````

Thank you for your response. `facet_wrap` might not be the best solution for me as I actually have many schools (70 schools) in my actual dataset. But second option is better.
I am actually doing this in a shiny app, with week appearing as a filter. When I select week 1, the graph appears correctly. But when I select week 1 and week 2, then the bars are shown wrong in the figure. Is there some way to overcome this?

This topic was automatically closed 21 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.