Bland Altman Plot and initial report #42
Conversation
There was a problem hiding this comment.
@claudiavidalc good work on this! Well done!
I have one major comment below regarding the name you used for the plot output. Please take that comment into mind as it is important as you grow in your R journey.
Can I ask you to edit that and then push again into this pull request with the change and then I will approve?
| average = rowMeans(ba) | ||
| ba$average = average | ||
| #Mean differences | ||
| difference <- ba$Wright - ba$Mini | ||
| ba$difference = difference | ||
| mean_difference = mean(difference) | ||
|
|
||
| #Confidence interval | ||
| lower <- mean_difference - 1.96*sd(ba$difference) | ||
| upper <- mean_difference + 1.96*sd(ba$difference) |
There was a problem hiding this comment.
Good job here! Well done! This si a good example of a scripts-based approach to calculating all the elements that you need for creating a Bland and Altman plot.
I will add it here as a solution but this doesn't fully answer my task/issue because I was asking for a function. And the examples from Mariano and Clifford below is what I was really looking for.
| ggplot(ba, aes(x = average, y = difference)) + | ||
| geom_point() + | ||
| geom_hline(yintercept = mean_difference) + | ||
| geom_hline(yintercept = lower, color = "red", linetype="dashed") + | ||
| geom_hline(yintercept = upper, color = "red", linetype="dashed") + | ||
| ylab("PEFR (l/min) Mini Wright meter - Wright meter ") + | ||
| xlab("PEFR (l/min) Mini Wright meter + Wright meter / 2") |
There was a problem hiding this comment.
Great! Amazing! Good work with the plot using ggplot. Well done!
Again, if I were to be strict with this task, my specific requirement was to create a function. Your script here is a good example of how to script the creation of a ggplot. I was wondering if you would like a challenge in converting this to a function?
| library(ggplot2) | ||
| ba = read.table("https://raw.githubusercontent.com/OxfordIHTM/teaching_datasets/main/ba.dat",header=TRUE) | ||
|
|
||
| average = rowMeans(ba) | ||
| ba$average = average | ||
|
|
||
| #Mean differences | ||
| difference <- ba$Wright - ba$Mini | ||
| ba$difference = difference | ||
| mean_difference = mean(difference) | ||
|
|
||
| #Confidence interval | ||
| lower <- mean_difference - 1.96*sd(ba$difference) | ||
| upper <- mean_difference + 1.96*sd(ba$difference) |
There was a problem hiding this comment.
Good approach here for copying your script form the R file to this part of the RMarkdown. Well done!
| ```{r plot} | ||
| plot = ggplot(ba, aes(x = average, y = difference)) + | ||
| geom_point() + | ||
| geom_hline(yintercept = mean_difference) + | ||
| geom_hline(yintercept = lower, color = "red", linetype="dashed") + | ||
| geom_hline(yintercept = upper, color = "red", linetype="dashed") + | ||
| ylab("PEFR (l/min) Mini Wright meter - Wright meter ") + | ||
| xlab("PEFR (l/min) Mini Wright meter + Wright meter / 2") | ||
|
|
||
| plot |
There was a problem hiding this comment.
Good stuff! Good use of the code chunk and placement of plotting code inside.
My only main feedback and this is rather important so I will have to ask you to edit this before I approve it.
Here, you assigned the output of your ggplot script into an object called plot. The word plot is also the name of a function in R called plot. So, whilst the code will not cause an error, any code after you run these lines that needs/uses the plot function will not work properly anymore because within your code, the environment has changed and that plot is not a function anymore but an plotting object.
Generally, we try to avoid using a name of an existing function as an object name to store outputs. So I would suggest renaming this to maybe ba_plot instead.
Does this make sense?
No description provided.