Plotting a graph where the majority of input data is very small but has large outliers

user_15601275 0
Viewed 31

So I have an app that takes in data from the user. 95% of the data will be within a very small range for example,0.411-0.412 depending on the user.

My problem is every so often there's data that will fall way outside that range for example 3 or 4, which is a problem when it comes to plotting the graph as it stretches to the maximum meaning you can't see the data as well.

For example the sample data [ 0.41,0.41,0.412,0.4112,0.415,0.415,0.4111,0.419,0.416] will produce: sample data 1

But if we add in a 4 to that data, the graph will produce:

Sample data 2

Whats a workaround to this? The y axis labels are split into 3 categories, cat1 cat2 cat3 depending on a reading, say 0.41 is cat1 0.42 is cat 3 and then anything above 0.44 is cat3.

I've tried capping the data by finding the average of the full set and changing any value that is 0.5 above the average to average+0.5. I'm not sure if thats the best way to do it as every user has different readings and 0.5 above the average may still be cat2 for some.

Any advice is appreciated.

graph math
1 Answers

Using a logarithmic scale is often a good approach for this. Each major division in your graph might be a doubling of the value, or a 10X increase. (The y scale becomes log₂(y) or log₁₀(y).)

user_2051850
Related