Question about APIC

Hi, I have a question about G2P in APIC while reading the given codes.
In the slides of G2P(APIC), Cp is updated with something like 4/dx^2. While in the given code “”, Cp is given by

new_C += 4 * weight * g_v.outer_product(dpos) * inv_dx

Only one “inv_dx” is presented here. I wonder why “inv_dx” is used here rather than “inv_dx**2”.
Thanks and looking forward to your reply.

Note that dpos already contains an inv_dx. If you compare dpos in p2g and g2p, you’ll find that they are different.

Got it. Thanks!