Box Plot Q3 Calculation

[Brian Abbott]

When working with a box plot or violin chart. The Q3 result is different to what I get in excel and chat gpt.
Using simple values as an example. 3 , 4 , 5 , 5 , 7, 8 , 9 , 10 , 11 , 11, 12.
Q3 in excel = 11 
Spotfire = 10.5 

from chat gpt
 

1. Median (Q2):

   • The dataset has an odd number of values, so the median is the middle value.
   • Median (Q2) = 8 (the 6th value in the sorted dataset)

2. First Quartile (Q1):

   • Counting from the start, Q1 is the median of the first half of the dataset.
   • Q1 = median of (3, 4, 5, 5, 7) = 5 (the 3rd value in the sorted dataset)

3. Third Quartile (Q3):

   • Counting from the end, Q3 is the median of the second half of the dataset.
   • Q3 = median of (9, 10, 11, 11, 12) = 11 (the 4th value in the sorted dataset)

So, for the given dataset, the median (Q2) is 8, the first quartile (Q1) is 5, and the third quartile (Q3) is 11.

 

[Andrew Berridge]

Hi Brian,

Great question!

As it happens, there are many different methods to calculate the quartiles.

Take a look at this Wikipedia page. There are at least 4 standard methods for calculating them:

https://en.wikipedia.org/wiki/Quartile

R can use one of 9 different methods...

Spotfire, and the Violin Mod use the type 7 method here:

https://www.rdocumentation.org/packages/stats/versions/3.6.2/topics/quantile

This is the default for R, and if it's good enough for R, I would say it's good enough for Spotfire.

However, if you absolutely NEED it to match Excel, then I could consider implementing the Excel-type calculations for the Violin Mod, as an option.

Let me know in a reply!

Thanks,

Andrew (Violin Mod author, Spotfire Field Tech Team)

 

 

[Brian Abbott]

Hello Andrew 

Thank you for the reply and for the link to the explanation. 

Thanks for the offer to alter the mod to match excel, however now that I have an understanding of how it is calculated it is all good as I will be able to explain it in the visualization.

Cheers 
Brian