Boundary scraping diagnostic (#3064)

* Initialize boundary scraping diagnostic

* Add new file

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Start fixing diagnostics

* Define ReadParameters function

* Define stubs for all required functions

* Enable new diagnostics in one of the tests

* Dump all particles at the end of the run

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Use the same types for different pinned-memory particles

* Use the same types for different pinned-memory particles

* Fix compilation error

* Finalize dumping buffer

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Clean up file

* Activate recording of particles at the boundary

* New function getParticleBufferPointer

* Pass a WarpXParticleContainer to the ParticleDiag

* Handle timestamp

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Revert to multi-level simulation

* Add documentation and a few checks

* Fix error with MPI

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Update varnames for RZ openPMD

* Add automated test

* Add automatic check

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Create new named component particle container

* Fix compilation

* Make WarpXParticleContainer derive from NamedParticleContainer

* Transfer functions that are specific to the particle component maps

* Move PIdx

* Make PinnedMemoryParticleContainer derive from NamedComponentParticleContainer

* Fix compilation

* Avoid unnecessary MPI communications

* Fix brittle code

* Define function `make_alike`

* Fix compilation

* Add back const anotations

* Fix typo

* Update code

* Fix compilation error

* Fix indentation

* Fix indentation

* Apply suggestions from code review

Co-authored-by: Axel Huebl <axel.huebl@plasma.ninja>

* Fix namespace

* Improve Error Messages Further

* Doxygen: Parameters of the Diag Constructor

* Update `WarpX-tests.ini`

... so we can filter the test out if we build w/o openPMD.

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Axel Huebl <axel.huebl@plasma.ninja>

1 files changed, 23 insertions, 0 deletions

diff --git a/Examples/Tests/scraping/analysis_rz.py b/Examples/Tests/scraping/analysis_rz.py
index 6c493fb0f..4af31e48a 100755
--- a/Examples/Tests/scraping/analysis_rz.py
+++ b/Examples/Tests/scraping/analysis_rz.py
@@ -13,6 +13,8 @@
 # Upon reaching the surface, particles should be removed.
 # At the end of the simulation, i.e., at time step 37,
 # there should be 512 particles left.
+# In addition, the test checks the boundary scraping diagnostic
+# by making sure that all removed particles are properly recorded.
 
 # Possible errors: 0
 # tolerance: 0
@@ -21,6 +23,8 @@
 import os
 import sys
 
+import numpy as np
+from openpmd_viewer import OpenPMDTimeSeries
 import yt
 
 sys.path.insert(1, '../../../../warpx/Regression/Checksum/')
@@ -38,5 +42,24 @@ print('error = ', error)
 print('tolerance = ', tolerance)
 assert(error==tolerance)
 
+# Check that all the removed particles are properly recorded
+# by making sure that, at each iteration, the sum of the number of
+# remaining particles and scraped particles is equal to the
+# original number of particles
+ts_full = OpenPMDTimeSeries('./diags/diag2/')
+ts_scraping = OpenPMDTimeSeries('./diags/diag3/')
+
+def n_remaining_particles( iteration ):
+    w, = ts_full.get_particle(['w'], iteration=iteration)
+    return len(w)
+def n_scraped_particles( iteration ):
+    timestamp = ts_scraping.get_particle( ['timestamp'] )
+    return (timestamp <= iteration).sum()
+n_remaining = np.array([ n_remaining_particles(iteration) for iteration in ts_full.iterations ])
+n_scraped = np.array([ n_scraped_particles(iteration) for iteration in ts_full.iterations ])
+n_total = n_remaining[0]
+assert np.all( n_scraped+n_remaining == n_total)
+
+# Checksum test
 test_name = os.path.split(os.getcwd())[1]
 checksumAPI.evaluate_checksum(test_name, fn, do_particles=False)