Timeline for Alternatives to three dimensional scatter plot
Current License: CC BY-SA 3.0
9 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 17, 2017 at 4:25 | comment | added | Tavrock |
The main difficulty I see is in having only integer values from 1 to 6. This makes it much harder to see what the data is doing as the points will overlap. You will most likely want to jitter your plotted points such as plot(jitter(y2) ~ jitter(x2), pch = 15) reference: thomasleeper.com/Rcourse/Tutorials/jitter.html
|
|
| Apr 15, 2017 at 20:18 | vote | accept | Ferdi | ||
| Apr 15, 2017 at 19:19 | answer | added | gung - Reinstate Monica | timeline score: 3 | |
| Apr 11, 2017 at 19:29 | history | tweeted | twitter.com/StackStats/status/851879813543493633 | ||
| Apr 11, 2017 at 13:52 | comment | added | G5W |
Parallel coordinate plots can be good at this scale (3 dimensions, 40 points) they are available through the parcoordfunction in the MASS package. Note that sometimes changing the order of the dimensions can make these plots more revelaing.
|
|
| Apr 11, 2017 at 13:33 | history | edited | Ferdi | CC BY-SA 3.0 |
added 106 characters in body
|
| Apr 11, 2017 at 13:29 | comment | added | SmallChess | In my field, the most example is the PCA plot. You lost only one dimension if you use PCA. | |
| Apr 11, 2017 at 13:27 | comment | added | Stephan Kolassa | Answers will depend on the structure and semantics of your data. Depending on what you have, you might use paneled scatterplots, or scatterplots with a third dimension indicated by colors. Can you tell us a little more about your data and maybe post a sample? | |
| Apr 11, 2017 at 13:25 | history | asked | Ferdi | CC BY-SA 3.0 |