Bug in how RPacketTracker tracks the rpacket.

[Sumit112192]

Describe the bug
Can someone please verify if the current implementation of RPacketTracker has a bug?
In the single_packet_loop function, the trace_packet function returns the distance, interaction_type, and delta_shell that the packet went through

tardis/tardis/transport/montecarlo/single_packet_loop.py

Lines 147 to 158 in 26b6005

	distance, interaction_type, delta_shell = trace_packet(
	r_packet,
	numba_radial_1d_geometry,
	time_explosion,
	opacity_state,
	estimators,
	chi_continuum,
	escat_prob,
	montecarlo_configuration.CONTINUUM_PROCESSES_ENABLED,
	montecarlo_configuration.ENABLE_FULL_RELATIVITY,
	montecarlo_configuration.DISABLE_LINE_SCATTERING,
	)

interaction_type could be BOUNDARY, LINE, ESCATTERING, etc.

tardis/tardis/transport/montecarlo/single_packet_loop.py

Lines 162 to 163 in 26b6005

	if interaction_type == InteractionType.BOUNDARY:
	move_r_packet(

tardis/tardis/transport/montecarlo/single_packet_loop.py

Lines 203 to 204 in 26b6005

	elif interaction_type == InteractionType.ESCATTERING:
	r_packet.last_interaction_type = 1

tardis/tardis/transport/montecarlo/single_packet_loop.py

Lines 230 to 231 in 26b6005

	elif (
	montecarlo_configuration.CONTINUUM_PROCESSES_ENABLED

Here's the issue
Say the packet went through a LINE interaction in shell_id=1; we track that in rpacket_tracker correctly, but now, if it went through a BOUNDARY, we are now using the previous value of r_packet to say that we have got a LINE interaction with the updated values of r and shell_id. Even though we are not going through a LINE interaction.Please correct me if I am wrong here.A proposed solution would be to update the rpacket_tracker only when it's not BOUNDARY Interaction.

Screenshots

As we can see, the last interaction type keeps getting copied over when the packet moves to the next boundary.

System

• OS:
  • GNU/Linux
  • macOS