Coordinate transform of discrete data

let's say I have a set of discrete data of dimension n by n by n, which may represent a scalar field $$f(x,y,z), x,y,z\in [-1/2,1/2]$$ in Cartesian 3D space. To be specific,

n=64;
data=Table[i+j+k,{i,-n/2,n/2},{j,-n/2,n/2},{k,-n/2,n/2}];


so that $$f(i/n,j/n,k/n)\simeq data[[i,j,k]]$$. My question is, is there an efficient way to calculate the "spherically integrated" data? i.e., in the continuous form I would like to calculate $$g(r)=\int f(r\sin\theta\cos\phi,r\sin\theta\sin\phi,r\cos\theta)r^2\sin\theta d\theta d\phi$$ In a less efficient way I would write something like

dataSph=Table[0,{l,0,2n}];
For[i=1,i<=n,i++,
For[j=1,j<=n,j++,
For[k=1,k<=n,k++,
For[l=1,l<=2n,l++,
If[l-1/2<=Sqrt[i^2+j^2+k^2]<l+1/2,dataSph[[l]]+=data[[i,j,k]]]
]]]]


or even just interpolate the data into a continuous version. Is there any better way to do the job?

• spherical integral can be write such as Integrate[Sqrt[x^2 + y^2 + z^2], Element[{x, y, z}, Sphere[{0, 0, 0}, r]]] Aug 24 '20 at 7:38
• You state that $(x,y,z) \in [-\frac{1}{2}, \frac{1}{2}]^3$ yet then compute Sqrt[i^2+j^2+k^2] to determine the "shells". Shouldn't there be an offset involved? E.g. Sqrt[(i-n/2)^2+(j-n/2)^2+(k-n/2)^2] or something similar. Aug 24 '20 at 16:41
• Yes there should be an offset. Thanks! Aug 25 '20 at 9:19

idx = Tuples[Range@n, 3];