Combine rotation commands

[ctk3b]

It might be more elegant to have just two rotation commands:

rotate(compound, thetas) and rotate_around_axis(compound, thetas)

where thetas is an array of 3 angles.

[sallai]

I had thought about that, too, but I decided not to. The problem is that
rotations around different axes are not commutable operations. It is a very
rare case that the user first wants to rotate around the x axis, then
around the y, and finally around the z.

On Tue, Dec 8, 2015 at 11:01 AM Christoph Klein notifications@github.com
wrote:

It might be more elegant to have just two rotation commands:

rotate(compound, thetas) and rotate_around_axis(compound, thetas)

where thetas is an array of 3 angles.

—
Reply to this email directly or view it on GitHub
#159.

[ctk3b]

Hmm yeah didn't quite think that through.

[chrisiacovella]

You can still use the rotate command to just rotate about one axis at a time if that were for some reason useful.

[sallai]

That is true. But still, I think that having the name of the method
explicitly describe what axis we want to rotate around would be a bit more
user friendly:

rotate_around_z(compound, np.pi/4)
rotate(compoud, [0, 0, np.pi/4])

On Wed, Dec 9, 2015 at 9:41 AM chrisiacovella notifications@github.com
wrote:

You can still use the rotate command to just rotate about one axis at a
time if that were for some reason useful.

—
Reply to this email directly or view it on GitHub
#159 (comment).

[ctk3b]

We may want to come up with a better name than revolve_around_x for rotating in place though since that sounds like it does the same thing. Any ideas?

I might even lean towards swapping the names or perhaps "spin"?

[chrisiacovella]

A spin operator sounds useful, since we often use that for initializing monolayers/bilayers (e.g., spin the molecule around it's long axis to randomize orientations). So to differentiate, spin could be a 2d operation, rotate is the 3d rotation matrix.

[sallai]

Yes, I like spin.

On Wed, Dec 9, 2015 at 1:01 PM chrisiacovella notifications@github.com
wrote:

A spin operator sounds useful, since we often use that for initializing
monolayers/bilayers (e.g., spin the molecule around it's long axis to
randomize orientations). I think that is better than calling it rotate,
since we can say rotate is the 3d rotation matrix.

—
Reply to this email directly or view it on GitHub
#159 (comment).

[chrisiacovella]

This might be useful for the forcefield fitting stuff, too. For instance, if we want to calculate the dihedral, we'll spin around a specific bond.