Remove NA out of plot

Hello! My situation is that I am able to run a set of code in R and produce plots using ggplot2 without specifying dropping N/A values. Its doing it in the background somehow. I am working on putting everything into a markdown file and at this particular set of code it isnt removing the n/a values for the data frame and producing the plots without n/a. In r markdown Im able to get plots but now it gives me a section in the bar chart with n/a. Any ideaswhy this is happening and how I can get r markdown to produce what I am getting in R?

This is the original dataframe with output below
dput(head(dailymerged, 5))

structure(list(Id = c(1503960366, 1503960366, 1503960366, 1503960366, 
1503960366), Date = structure(c(16903, 16904, 16906, 16907, 16908
), class = "Date"), Weekday = structure(c(3L, 4L, 6L, 7L, 1L), .Label = c("Sun", 
"Mon", "Tue", "Wed", "Thu", "Fri", "Sat"), class = c("ordered", 
"factor")), Calories = c(1985L, 1797L, 1745L, 1863L, 1728L), 
    TotalSteps = c(13162L, 10735L, 9762L, 12669L, 9705L), TotalDistance = c(8.5, 
    6.96999979019165, 6.28000020980835, 8.15999984741211, 6.48000001907349
    ), VeryActiveDistance = c(1.87999999523163, 1.57000005245209, 
    2.14000010490417, 2.71000003814697, 3.19000005722046), ModeratelyActiveDistance = c(0.550000011920929, 
    0.689999997615814, 1.25999999046326, 0.409999996423721, 0.779999971389771
    ), LightlyActiveDistance = c(6.05999994277954, 4.71000003814697, 
    2.82999992370605, 5.03999996185303, 2.50999999046326), SedentaryDistance = c(0, 
    0, 0, 0, 0), VeryActiveMinutes = c(25L, 21L, 29L, 36L, 38L
    ), ModeratelyActiveMinutes = c(13L, 19L, 34L, 10L, 20L), 
    LightlyActiveMinutes = c(328L, 217L, 209L, 221L, 164L), SedentaryMinutes = c(728L, 
    776L, 726L, 773L, 539L), TotalSleepRecords = c(1L, 2L, 1L, 
    2L, 1L), TotalMinutesAsleep = c(327L, 384L, 412L, 340L, 700L
    ), TotalTimeInBed = c(346L, 407L, 442L, 367L, 712L)), row.names = c(NA, 
5L), class = "data.frame")

From dailymerged I created another dataframe named ActivityLong.

structure(list(Id = c(1503960366, 1503960366, 1503960366, 1503960366, 
1503960366), Date = structure(c(16903, 16903, 16903, 16903, 16904
), class = "Date"), Weekday = structure(c(3L, 3L, 3L, 3L, 4L), .Label = c("Sun", 
"Mon", "Tue", "Wed", "Thu", "Fri", "Sat"), class = c("ordered", 
"factor")), Calories = c(1985L, 1985L, 1985L, 1985L, 1797L), 
    TotalSteps = c(13162L, 13162L, 13162L, 13162L, 10735L), TotalDistance = c(8.5, 
    8.5, 8.5, 8.5, 6.96999979019165), Level = structure(c(4L, 
    3L, 2L, 1L, 4L), .Label = c("Sedentary", "LightlyActive", 
    "ModeratelyActive", "VeryActive"), class = "factor"), Distance = c(1.87999999523163, 
    0.550000011920929, 6.05999994277954, 0, 1.57000005245209), 
    Minutes = c(25, 13, 328, 728, 21), TotalSleepRecords = c(1L, 
    1L, 1L, 1L, 2L), TotalMinutesAsleep = c(327L, 327L, 327L, 
    327L, 384L), TotalTimeInBed = c(346L, 346L, 346L, 346L, 407L
    )), row.names = c(NA, -5L), class = c("tbl_df", "tbl", "data.frame"

Here is the code I run to get ActivityLong

dailymerged <- dailymerged |> 

ActivityLong <- dailymerged |> 
  pivot_longer(VeryActiveDistance:SedentaryMinutes, names_to = "Level_Metric")

head(ActivityLong, 3)

Running this in R ends up giving me ActivityLong with 1640 obs of 12 variables and then seems to be using that to make the plots below. But, in R markdown I cant get the dataframe to automatically drop the n/a values and its giving me 2-3000 obs and then making the plots which produce a warning of removing X # of row with n/a values but then the pplots have the n/a values in them. See below.

This is the code I am running to get plots in R.

ggplot(ActivityLong, aes(x= Weekday, y= Minutes, fill=Level)) + 

ggplot(ActivityLong, aes(x = Weekday, y = Minutes, fill=Level)) + 

The plot in the middle is what I am getting in R markdown and the plot to the R is what I am getting in R. How do I get R markdown to produce the same plot as in R?


The following example uses the starwars dataset and applies the na.omit() function to remove NA results in the dataset. Try this with your dataset and see what happens.

starwars %>%
select(name, gender, hair_color, height) %>%
na.omit() %>% view()


I logged in and attempted to do what you suggested and it didnt change anything. I ended up finding myself in an endless loop of being told that the packages loaded needed to be updated and that I needed to restart. So I would and then I tried reloading the packages and would get the same thing. This process is very frustrating to have it work one day and not the next. To have it work in one format and not the next. I really feel like Im wasting a lot of time and going nowhere because of endless errors.

I can relate to your situation and feel for you, it sucks to waste so much time and get nowhere. I've just spent about a week trying to change a variable name in a dataset for an assignment - sometimes I felt like headbutting the computer and screaming. Hope an answer comes along soon.

Hmm, that is strange that it is showing the NAs in the Markdown. One thing that may help is to pipe in the ggplot statement in RMarkdown as part of your process flow.

The following code (in theory) is a similar thought to what MrX recommended, where you can filter the data further after you are finished cleaning it and then plot in one code chunk:


ActivityLong %>% # %>% is equivalent to |>
  filter (! %>% 
  ggplot(aes(x= Weekday, y= Minutes, fill=Level)) + 

All the Level NAs should be scrubbed prior to plotting. Alternatively, you can filter the Level variable to just be a valid value aka = c("Sedentary", "LightlyActive", "ModeratelyActive", "VeryActive").

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.