Synching ALERT fork with main branch #378

mpaolone · 2024-12-02T15:38:25Z

No description provided.

It should have been AHDC::adc (since corresponding HitReader also uses this bank), it was AHDC::tdc.

…pment

…om TDC to ADC.

(and stop changing wire number internally to start from zero) (this requires removing a manual +1 offset for wire number in a perl script in GEMC)

Iss272 alert indexing

get latest changes

Fix of a merge error after commit f10c09e

PreCluster pairing routine in AHDC ClusterFinder

…r-layer list of hits in AHDC PreCluster Modified the hardcoded lookup superlayer indices to obtain the geometry parameters to calculate x, y.

Modifications for debugging/validation of index convention change.

…ector_simple" for Hit, and "H_simple", "h_simple" for KFitter to return the preexisting matrix emasurements; new matrix measurements have been implemented in the methods "getMeasurementNoise", "getVector", "H" and "h". Added new memebers "adc" and "phi" for the hit for measured ADC and wire phi position at z = 0 respectively.

Getting changes from Michael

…pment

common-tools/clas-detector/src/main/java/org/jlab/detector/base/DetectorDescriptor.java

raffaelladevita · 2024-12-05T15:51:24Z

reconstruction/alert/src/main/java/org/jlab/rec/ahdc/Cluster/Cluster.java

@@ -27,7 +27,9 @@ public Cluster(PreCluster precluster, PreCluster other_precluster) {
 		_PreClusters_list.add(precluster);
 		_PreClusters_list.add(other_precluster);
 		this._Radius = (precluster.get_Radius() + other_precluster.get_Radius()) / 2;
-		this._Z      = ((other_precluster.get_Phi() - precluster.get_Phi()) / (Math.toRadians(20) * Math.pow(-1, precluster.get_Super_layer()) - Math.toRadians(20) * Math.pow(-1, other_precluster.get_Super_layer()))) * 300 - 150;
+
+		this._Z      = ((other_precluster.get_Phi() - precluster.get_Phi()) / (Math.toRadians(20) * Math.pow(-1, precluster.get_Super_layer()-1) - Math.toRadians(20) * Math.pow(-1, other_precluster.get_Super_layer()-1))) * 300 - 150;


hardcoded numbers should be replaced with variables possibly defined from the database or in this case the geometry service

raffaelladevita · 2024-12-05T15:52:46Z

reconstruction/alert/src/main/java/org/jlab/rec/ahdc/Cluster/ClusterFinder.java

 		if (precluster.get_Super_layer() == super_layer && precluster.get_Layer() == layer && !precluster.is_Used()) {
 			ArrayList<PreCluster> possible_precluster_list = new ArrayList<>();

-			double phi_mean = precluster.get_Phi() + 0.1 * Math.pow(-1, precluster.get_Super_layer());
+			double phi_mean = precluster.get_Phi() + 0.1 * Math.pow(-1, precluster.get_Super_layer()-1);


same comment about hardcoded numbers

I think it's the "0.1" in question

raffaelladevita · 2024-12-05T15:54:00Z

reconstruction/alert/src/main/java/org/jlab/rec/ahdc/Hit/Hit.java

 		switch (this.superLayerId) {
-			case 0:
+			case 1:


same comment about hardcoded numbers

reconstruction/alert/src/main/java/org/jlab/rec/service/AHDCEngine.java

raffaelladevita · 2024-12-05T17:48:39Z

etc/bankdefs/hipo4/data.json

comments in the ATOF::add bank should be updated to be consistent with sector/layer starting from 1

etc/bankdefs/hipo4/data.json

raffaelladevita · 2024-12-05T17:58:42Z

reconstruction/alert/src/main/java/org/jlab/rec/ahdc/PreCluster/PreClusterFinder.java

@@ -23,21 +23,21 @@ private void fill_list(List<Hit> AHDC_hits, ArrayList<Hit> sxlx, int super_layer

 	public void findPreCluster(List<Hit> AHDC_hits) {
 		ArrayList<Hit> s0l0 = new ArrayList<Hit>();
-		fill_list(AHDC_hits, s0l0, 0, 0);
+		fill_list(AHDC_hits, s0l0, 1, 1);


The List of Lists should be filled all at once to avoid looping 12 times through the hits

raffaelladevita · 2024-12-05T18:13:58Z

common-tools/clas-detector/src/main/java/org/jlab/detector/pulse/Pulse.java

+    public int id;
+
+    public float adcMax;
+    public float timeRiseCFA;


Since this class is for general use, members' names should be more generic removing the CFA or CFD suffix.

Also the pulse "rise" and "fall" are usually referred to as leading and trailing edge that could be better names

raffaelladevita · 2024-12-05T18:24:38Z

common-tools/clas-detector/src/main/java/org/jlab/detector/pulse/ModeAHDC.java

+	public long timeStamp;
+	public float fineTimeStampResolution;
+	public static final short ADC_LIMIT = 4095; // 2^12-1
+	public float amplitudeFractionCFA;


All the pulse specific variables should be local variables in the extract method

yes, good catch, critical for thread safety

…tem, one to minimize track distance. Added functions to calculate phi for hit as a function of z and to calculate the hit "sign" i.e. the side of the wire the track passes.

…ers all hits to make a track candidate and uncommented the track finding code.

Kalman Filter fixes: constants and parameters

baltzell · 2024-12-10T22:36:27Z

And the creation of BMT::wf (and conversion to BMT::adc) in the decoder was intended only as a working example, to be replaced with AHDC.

baltzell · 2024-12-10T23:00:54Z

Ok, after this is merged, then the Mode3 class in the decoder needs to be switched to the one AHDC wants to use.

fix bad merge conflict

…r to not loop 12 more times than needed

baltzell · 2024-12-13T17:54:05Z

And this will address that and that (and create the still missing AHDC::wf bank definition)

Actually use the new waveform conversion approach for AHDC

Use DeepJavaLibrary to load a PyTorch model. Create a new bank AHDC_AI::Prediction.

whit2333

ATOF geometry should be untouched.

common-tools/clas-geometry/src/main/java/org/jlab/geom/detector/alert/ATOF/AlertTOFFactory.java

This reverts commit 3420bb9.

baltzell · 2024-12-16T18:49:02Z

This is now required.

Revert "leave internal indexing at zero, output at 1"

Add AI for ALERT trackfinding

baltzell · 2024-12-18T20:56:15Z

Who can resolve the ATOF geometry conflict?

efuchey and others added 30 commits February 15, 2024 20:10

Changed the data bank to process the hits in AHDCEngine:

87011e4

It should have been AHDC::adc (since corresponding HitReader also uses this bank), it was AHDC::tdc.

Merge branch 'JeffersonLab:development' into development

3863e74

reversed the change in AHDCEngine.

10559cd

Merge branch 'development' of github.com:efuchey/coatjava into develo…

8ac4238

…pment

Merge branch 'development' of github.com:efuchey/coatjava into develo…

e9377c4

…pment

Merge branch 'JeffersonLab:development' into development

81b637e

Merge branch 'JeffersonLab:development' into development

4046143

Reverted (again???) the hit type to be read in AHDC reconstruction fr…

08cd675

…om TDC to ADC.

changed AHDC bank back to ADC only

10f0972

Merge branch 'JeffersonLab:development' into development

acde32e

remove redundant checks

9440e21

remove unused objects

c95e068

cleanup

11d6954

leave internal indexing at zero, output at 1

3420bb9

cleanup

d6fd779

remove redundant checks

46e33d0

leave internal indexing from zero, output from 1

268c239

simplicy logic

4a9647f

use s/l/c/o indexing starting from 1

481795d

(and stop changing wire number internally to start from zero) (this requires removing a manual +1 offset for wire number in a perl script in GEMC)

Merge pull request #1 from JeffersonLab/iss272-alert-indexing

04571a1

Iss272 alert indexing

merge post Nathan's indexing update

082ae77

Merge branch 'JeffersonLab:development' into development

f137c4f

Merge branch 'JeffersonLab:development' into development

7c6593a

Merge pull request #1 from mpaolone/development

f10c09e

get latest changes

fixed a merge error that slipped through the cracks.

2e86a66

Merge branch 'JeffersonLab:development' into development

7514d48

Merge pull request #2 from efuchey/development

5ae48d4

Fix of a merge error after commit f10c09e

Modified the reference layer index for the "OtherCluster" in the

85dbe45

PreCluster pairing routine in AHDC ClusterFinder

Modified the hardcoded lookup superlayer/layer indices to fill the pe…

e14c464

…r-layer list of hits in AHDC PreCluster Modified the hardcoded lookup superlayer indices to obtain the geometry parameters to calculate x, y.

Merge pull request #4 from efuchey/development

4b3603b

Modifications for debugging/validation of index convention change.

efuchey and others added 3 commits December 3, 2024 13:54

Merge pull request #4 from mpaolone/development

7fe0c41

Getting changes from Michael

Merge branch 'development' of github.com:efuchey/coatjava into develo…

f379279

…pment

baltzell assigned raffaelladevita Dec 5, 2024

raffaelladevita requested changes Dec 5, 2024

View reviewed changes

baltzell mentioned this pull request Dec 6, 2024

The waveform proposal #352

Closed

efuchey and others added 3 commits December 9, 2024 10:34

Duplicated the measurement matrices functions: one in cylindrical sys…

5dd9fe4

…tem, one to minimize track distance. Added functions to calculate phi for hit as a function of z and to calculate the hit "sign" i.e. the side of the wire the track passes.

For pull request only: commented out the line in AHDCEngine that gath…

0723c15

…ers all hits to make a track candidate and uncommented the track finding code.

Merge pull request #22 from efuchey/development

506e888

Kalman Filter fixes: constants and parameters

baltzell and others added 2 commits December 10, 2024 17:41

Merge branch 'development' into development

9abc31a

fix bad merge conflict

3867efe

baltzell mentioned this pull request Dec 10, 2024

fix bad merge conflict mpaolone/coatjava#23

Merged

mpaolone and others added 4 commits December 12, 2024 11:12

Merge pull request #23 from baltzell/development

a25a4d1

fix bad merge conflict

changed ModeAHDC to be dumber and threadsafe, changed PreClusterFinde…

5636a1a

…r to not loop 12 more times than needed

AHDC::wf must be defined

d6adcd8

actually use the new ahdc pulse extractor

0bd4108

mpaolone and others added 2 commits December 16, 2024 08:59

Merge pull request #24 from baltzell/development

48eaa66

Actually use the new waveform conversion approach for AHDC

Add AI for ALERT trackfinding

3521313

Use DeepJavaLibrary to load a PyTorch model. Create a new bank AHDC_AI::Prediction.

whit2333 requested changes Dec 16, 2024

View reviewed changes

common-tools/clas-geometry/src/main/java/org/jlab/geom/detector/alert/ATOF/AlertTOFFactory.java Outdated Show resolved Hide resolved

Revert "leave internal indexing at zero, output at 1"

bd001db

This reverts commit 3420bb9.

baltzell requested review from whit2333, raffaelladevita and baltzell December 16, 2024 18:50

mpaolone added 2 commits December 18, 2024 09:35

Merge pull request #26 from baltzell/development

e2acd07

Revert "leave internal indexing at zero, output at 1"

Merge pull request #25 from mathieuouillon/AI_for_ALERT

88b1dd8

Add AI for ALERT trackfinding

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Synching ALERT fork with main branch #378

Synching ALERT fork with main branch #378

mpaolone commented Dec 2, 2024

raffaelladevita Dec 5, 2024

raffaelladevita Dec 5, 2024

baltzell Dec 6, 2024

raffaelladevita Dec 5, 2024

raffaelladevita Dec 5, 2024

raffaelladevita Dec 5, 2024

raffaelladevita Dec 5, 2024

raffaelladevita Dec 5, 2024

baltzell Dec 6, 2024

baltzell commented Dec 10, 2024 •

edited

Loading

baltzell commented Dec 10, 2024

baltzell commented Dec 13, 2024 •

edited

Loading

whit2333 left a comment

baltzell commented Dec 16, 2024

baltzell commented Dec 18, 2024

Synching ALERT fork with main branch #378

Are you sure you want to change the base?

Synching ALERT fork with main branch #378

Conversation

mpaolone commented Dec 2, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

baltzell commented Dec 10, 2024 • edited Loading

baltzell commented Dec 10, 2024

baltzell commented Dec 13, 2024 • edited Loading

whit2333 left a comment

Choose a reason for hiding this comment

baltzell commented Dec 16, 2024

baltzell commented Dec 18, 2024

baltzell commented Dec 10, 2024 •

edited

Loading

baltzell commented Dec 13, 2024 •

edited

Loading