Matrices result in data points, and a surface plot is based on the grid of the data. Vectors result in a scatter plot, and a surface plot is generated by interpolating a grid. The graph seems to decide whether the basis is a scatter plot or data points depending on whether the input arrays are vectors or matrices. But the outer edges, which are extrapolated rather than interpolated, are still considerably off. If you minimize the interpolation by setting the interpolated mesh to match your data (26 rows, 7 columns) the plots look a lot better. Your data appears to be very close to a regular grid. It appears to be a problem specifically with the plot's handling of the interpolation, as the standard cubic spline interpolation, either one dimensional or two dimensional, does not exhibit this problem. That case show some problems at the edges of the plot. I worked with a simpler case, both x and y runing as integers from -3 to 3, with the same boustrophedoninc order. Your data is fairly similar to what I surmised, so what I found is applicable. Can you post the data aand worksheet? I am curious to see how Axum handles it.Īpparently the data table component was changed for MC11, and so such tables from MC11 don't work in 2001i. You should add a description like this to the "contour, surface, scatter plot & pictures" section of the data analysis pack. The 2D plot at right shows the distribution of the (x,y) locations. On the same 3D plot I use a scatter and a surface plot of some (x,y,z) data. Now, I've exhumed some old data that shows that even in the case where you try to control the 3D plot by providing three 2D arrays for X, Y and Z you can still obtain obvious mistakes in the surface generation. It seems that in that case the plot routine switches to another mode where there can be connectivity between any of the input data points. In your original post you passed 3 vectors of data, which define a 1D ribbon. That's why in the plot I showed earlier you can have a U-shaped surface instead of a filled square surface. However, that's the limit of the continuity. The plot routine will create surface elements that span the gap between successive segments. The ribbon is made of a succession of such segments. Each row in the X, Y and Z matrices defines the position of a line segment. One way of controlling the 3D plot is to think of the surface as a ribbon.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |