JMP可否做出3D的Contour Plot？

newcop

我有一個2X2X2 Full Factorial的試驗，Contour Plot可以針對任意二個變數X做出Response的圖。

那我想請問一下，JMP可否同時使用三個變數，針對一個Response做出3D的Contour Plot呢？

lsc771017

可以的，没问题的

newcop

太好了，可以請問一下該從哪邊操作呢？
我從Graph→Contour Plot，裡面我最多只能選擇2個X，X3就不讓我加入了

newcop

自己回覆一下現在的狀況
在SAS上詢問的結果是這樣：
JMP does not have a 3-d plot that will allow the addition of contours. However, the "Graph->Contour Profiler" platform will allow you to get contours and it offers slider bars to move the value of factors not being plotted (X3, for example). Similarly, the "Graph->Surface Plot" platform will give a 3-dimensional view of the surface and the slider bar for "X3" is also present.   

To illustrate, open the attached data table and select "Graph->Contour Profiler".  Enter "Pred Formula Y" as the "Y, Prediction Formula".  Click "Ok".  From the inverted red arrow next to "Contour Profiler", select "Contour Grid".  Define the grid values, and click "Ok".  Now, you can move the slider bar for "X3" to see the changing contours across the values of X3.

Or, to see the surface plot, select "Graph->Surface Plot".  Enter "Pred Formula Y" in the "Columns" role.  Click "Ok".  Here, you can use the slider bar for "X3" to see the surface at different values of X3.   

不知道是不是我表達得不好？SAS的回覆是無法做出3D的Contour Plot？只能用像附件那樣的方式去看X3的值。

但是這樣就無法得到X1、X2、X3三因子變動範圍對應到Response Y的等高線圖了。

請問一下 lsc771017 跟其他高手，是否真有辦法做出3D的Contour Plot呢？

hogan

3因子的用Prediction Profiler 就可以了，效果比Contour profiler好很多.