Hi Kevin,

Let me explain it with an example.

For example, one-dimensional heat equation.
u_t-u_xx=f
If I discrete time and space in t_0, t_1,...,t_m, and x_0,x_1,...,x_n
respectively. We want to store these points. Then it is a matrix u(t_i,
x_j).

What I am concern is how to store such a big matrix. When we solve this
equation, it is often the case that we get at first the u(t_0, *)s, and
then u(t_1, *)s, and then u(t_2, *), and so on and so forth.

For two or three dimensional heat equation, we want a multi-dimensional
array e.g. u(t_i, x_j, y_k). We may sometimes use u(t_i, z_k) instead, in
which k = 1,2,...,m*n, that is to say, we use a long array to replace the
last two sub-indexes in matrix u(t_i, x_j, y_k).

So what is a good strategy to create and store such "matrices"?

Yi


On Mon, May 26, 2014 at 12:15 PM, Kevin Squire <[email protected]>wrote:

> Hello Yi,
>
> Your question is rather open-ended.  Can you give an example of the kind
> of multi-dimensional data you're interested in, and what you want to
> visualize about it?
>
> Cheers,
>    Kevin
>
>
> On Sun, May 25, 2014 at 6:40 PM, yi lu <[email protected]>wrote:
>
>> In scientific computing, we need multi-dimensional arrays to *store*time and 
>> space grid points for computing and for plotting. So, what is best
>> *recommended* to use in Julia for such requirement? What I mean by best
>> is fastest, cheapest and so on.
>>
>> Thanks in advance.
>>
>> Yi
>>
>
>

Reply via email to