In line 749 of transform3d.py,
It keeps causing an error when I try to visualize sdf using pv.draw_sdf_slice because a is a float tensor while b is a double tensor.
Just a heads up since I'm not sure how critical this is when using other functions in pytorch_kinematics.
I faced this issue when I tried to specify dtype of a chain class.