sakurako97
New Member
- Joined
- Nov 19, 2012
- Messages
- 3
Hi,
I have a dataset which is widely distributed. A log chart can display the pos values fine, but I need the same treatment for neg values also which obviously rules out log.
My values for example range from -5 to +5 for both axes.
There are a lot of data points in the 0 to .03 range and -.03 to 0 range, but there is also a much smaller, yet fairly well distributed set of values between 1 and 5(both pos and neg). If I use a normal linear scale, it just bunches all the points, and to make it look even semi readable, it would need to be a very large graph.
I don;t want to transform the data as the numbers have particular meaning in terms of thresholds etc, and it would be confusing to the reader.
Any ideas would be much appreciated.
Thank you
I have a dataset which is widely distributed. A log chart can display the pos values fine, but I need the same treatment for neg values also which obviously rules out log.
My values for example range from -5 to +5 for both axes.
There are a lot of data points in the 0 to .03 range and -.03 to 0 range, but there is also a much smaller, yet fairly well distributed set of values between 1 and 5(both pos and neg). If I use a normal linear scale, it just bunches all the points, and to make it look even semi readable, it would need to be a very large graph.
I don;t want to transform the data as the numbers have particular meaning in terms of thresholds etc, and it would be confusing to the reader.
Any ideas would be much appreciated.
Thank you