[ybedfer]

Briefly, what does this PR introduce? Please link to any relevant presentations or discussions.

Allows to run eicrecon on 12-Sector-CyMBaL (see PR#1146 of epic: "eic/epic#1146")

What is the urgency of this PR?

• [* ] Medium

What kind of change does this PR introduce?

• [ *] other:

Please check if any of the following apply

• No change to the default behaviour.

[ybedfer]

When the current material map is used, eicrecon behaves badly on 12-Sector-CyMBaL: no track reconstructed. The following error message is emitted in the initialization step:

[] [acts] [error] Error during DD4Hep -> ACTS geometry conversion: Surface::assignSurfaceMaterial: material axis directions {AxisPhi, AxisZ} are not supported by this surface. Supported axes are: {AxisR, AxisPhi}

This, since acts,v46 has been in use.

Here, we add a dedicated material-map_CyMBaL_12S.cbor map.

• obtained w/ ./run_material_map_validation.sh --nevents 8000 --nparticles 5000,
• checked to fix the above-mentioned misbehavior and error,
• yielding the following validation plots:

=> Somewhat imperfect @ large >0 η...

...Yet good enough to allow the review of epic's PR #1146.

Note: see infra for a picture of the validation ration as a f(η).

[ybedfer]

Ratio_Val_geant_mat_X0_Eta.pdf

[veprbl]

So what is the plan to enable this? Are you going to switch to 2D strip by default in eic/epic#1146 ? Or we need to switch material map based on the configuration?

[ybedfer]

So what is the plan to enable this? Are you going to switch to 2D strip by default in eic/epic#1146 ? Or we need to switch material map based on the configuration?

I had been expecting you to tell me...

At the last tracking meeting, as far as I understand, it was agreed that 12-Sector was second order of priority.
If this is correct, then your option #2 is the answer. And then, question: how can this (swicth material map) be done?

[ShujieL]

So what is the plan to enable this? Are you going to switch to 2D strip by default in eic/epic#1146 ? Or we need to switch material map based on the configuration?

I had been expecting you to tell me...

At the last tracking meeting, as far as I understand, it was agreed that 12-Sector was second order of priority. If this is correct, then your option #2 is the answer. And then, question: how can this (swicth material map) be done?

The material map is provided at https://github.com/eic/epic/blob/abb5202d647b558018335549eedefae0674d9263/compact/tracking/definitions_craterlake.xml#L218. So if we want to support two versions of geometry, we can have two versions of the definition files e.g. definitions_craterlake_mpgd.xml for https://github.com/eic/epic/blob/main/configurations/craterlake_tracking_2DStrip.yml

[ybedfer]

@ShujieL
Hello,
I have modfied my definitions_craterlake.xml:

diff --git a/compact/tracking/definitions_craterlake.xml b/compact/tracking/definitions_craterlake.xml
index 4679216db..9421a3be9 100644
--- a/compact/tracking/definitions_craterlake.xml
+++ b/compact/tracking/definitions_craterlake.xml
@@ -217,2 +217,2 @@
-      <arg value="file:calibrations/materials-map.cbor"/>
-      <arg value="url:https://raw.githubusercontent.com/eic/epic-data/700fa7f92568311408137b2002b513346b66277d/material-map.cbor"/>
+      <arg value="file:calibrations/material-map_CyMBaL_12S.cbor"/>
+      <arg value="url:https://raw.githubusercontent.com/eic/epic-data/2e030ad8a682f36db05e72d511c727fc95b1071c/material_CyMBaL_12S.cbor-map.cbor"/>

where the SHA1 is that if the HEAD of epic-data:

$ git rev-parse HEAD
2e030ad8a682f36db05e72d511c727fc95b1071c

Then I get the following error from eicrecon:

FileLoader       ERROR unable to run the download command curl --retry 5 --location --fail https://raw.githubusercontent.com/eic/epic-data/2e030ad8a682f36db05e72d511c727fc95b1071c/material_CyMBaL_12S.cbor-map.cbor --output calibrations/fac8c5c3e6a03fe6

What did I miss? How can I check the above-mentioned url raw.githubusercontent.com/eic/epic-data

[veprbl]

You have a mistake here: material_CyMBaL_12S.cbor-map.cbor

[ybedfer]

You have a mistake here: material_CyMBaL_12S.cbor-map.cbor

Thanks!
Stupid mistake.
But it still doesn't work.
I'm investigating...

...When I do not specify any -Pacts:MaterialMap option on the eicrecon command line (w/ an argument pointing to my local epic_data), acts tries to load the material map w/ the usual filename:

[] [acts] [info] loading materials map from file: 'calibrations/materials-map.cbor'

File, which it is not able to find later on:

Unable to open input JSON material file: calibrations/materials-map.cbor

given that eicrecon has retrieved, and stored in the calibrations directory, a material-map_CyMBaL_12S.cbor file.

My understanding is that eicrecon does get the info use-alternative-material-map, thanks to the <plugin> tag in definitions_craterlake.xml, but fails to convey it over to acts.

[ybedfer]

Hello again,
It works when I supply the filename of the map w/ the option -Pacts:MaterialMap and tell acts to get it from the calibrations sub-directory:

-Pacts:MaterialMap=calibrations/material-map_CyMBaL_12S.cbor

[ybedfer]

I finally understood how the whole thing is working:

• Filename has to keep being calibrations/materials-map.cbor.
• And path has to point to the location of the file on the repository, whatever is name.

[veprbl]

-Pacts:MaterialMap option should work, if not file an issue.

[ybedfer]

The option is not necessary. I am to commit a new version of definitions_craterlake.xml (in my 12-Sector-CyMBaL branch) to that end (once all my checks are completed).

[veprbl]

What might be related eic/epic#494 if we want to setup custom file names.

[ybedfer]

I have just pushed a new version of my 12-Sector-CyMBaL branch of epic.
The github checks for the epic PR (#1146: "eic/epic#1146") should work. Except for those involving 2DStrip, which in addition require the EICrecon PR (#2829: "eic/EICrecon#2829".