l******n
发帖数: 9344 | 1 我有很多的数据，是在一个球面上的概率分布函数，但是这些数据的坐标在纬度不
是equally spaced的，（经度是），而且在2极是附近时没有的
现在做图，希望能够用matlab的函数能够画出完整的图
有人知道那个函数比较？而且怎么做比较好看？
写了 | r****y
发帖数: 1437 | 2 For each meridional circle, do interpolation to regular lat grid
then any map toolbox function, like surfm, contourm, can do the job.
【在 l******n 的大作中提到】
: 我有很多的数据，是在一个球面上的概率分布函数，但是这些数据的坐标在纬度不
: 是equally spaced的，（经度是），而且在2极是附近时没有的
: 现在做图，希望能够用matlab的函数能够画出完整的图
: 有人知道那个函数比较？而且怎么做比较好看？
: 写了
| l******n
发帖数: 9344 | 3 I only want to make up the pole points to make it look like a whole sphe
re, and I do not want do interpolation to regular lat.
I am asking if there is a function in matlab can do this?
I think there is one, but could not remember
thanks
job.
【在 r****y 的大作中提到】
: For each meridional circle, do interpolation to regular lat grid
: then any map toolbox function, like surfm, contourm, can do the job.
| r****y
发帖数: 1437 | 4 even you do not do interpolation, any plot command has to do it
by all means.
If you have no data at pole, how could you make it up?
【在 l******n 的大作中提到】
: I only want to make up the pole points to make it look like a whole sphe
: re, and I do not want do interpolation to regular lat.
: I am asking if there is a function in matlab can do this?
: I think there is one, but could not remember
: thanks
:
: job.
| l******n
发帖数: 9344 | 5 that is why I ask for command which will do interpolation automatically
【在 r****y 的大作中提到】
: even you do not do interpolation, any plot command has to do it
: by all means.
: If you have no data at pole, how could you make it up?
| r****y
发帖数: 1437 | 6 it just needs <=3 lines of code. If you cannot do this without
a command, no future la.
【在 l******n 的大作中提到】
: that is why I ask for command which will do interpolation automatically
| l******n
发帖数: 9344 | 7 the problem is I did as you said, but the interpolation give NaN values
to me.
I am little confused by this.
The function I am computing is smooth on the sphere under the sphere coo
rdinates, it might be singular to (x,y,z) coordinates.
【在 r****y 的大作中提到】
: it just needs <=3 lines of code. If you cannot do this without
: a command, no future la.
| r****y
发帖数: 1437 | 8 from what you said, you have to do interpolation for the whole
circle, not just one meridional curve, the half circle.
If you have no data above certain latitude band, no need to bother,
you are faking data in this sense. hoho.
【在 l******n 的大作中提到】
: the problem is I did as you said, but the interpolation give NaN values
: to me.
: I am little confused by this.
: The function I am computing is smooth on the sphere under the sphere coo
: rdinates, it might be singular to (x,y,z) coordinates.
| l******n
发帖数: 9344 | 9 I did do whole circle interpolation.
Let me make it more clear.
suppose you are given a N by M matrix F, which is the values of the dens
ity function on the sphere, the corresponding coordinates are
x(i,j) = cos(theta(i))*cos(phi(j))
y(i,j) = cos(theta(i))*sin(phi(j))
z(i,j) = sin(theta(i))
I can plot the density on the sphere no problem, but as I said there are
holes in poles due to theta phi values.
For display purpose, I want to fix the wholes.
I want to change theta to THETA, which is equall
【在 r****y 的大作中提到】
: from what you said, you have to do interpolation for the whole
: circle, not just one meridional curve, the half circle.
: If you have no data above certain latitude band, no need to bother,
: you are faking data in this sense. hoho.
|
|
