Add checksum regression tests (#1065)

* add a checksum module for regtests

* All capabilities to checksum tests

* update tests to generate a clean benchmark

* call checksum tests from analysis scripts

* Add checksum json benchmarks

* add missing script, better printing and increase tol

* eol

* fix permission, and bump tolerance

* chmod +x checksumAPI.py

* update benchmarks, probably old ones were wrong due to parallelization

* remove unused variable, thanks lgtm

* special handling when the benchmark is 0

* non-zero but very small CFL for 1 step for init tests

* reset erroneous Python benchmarks

* slightly bump tolerance for initial distribution test

* too small dr results in nans

* improve prints, tolerance handling, and fix tests

* typo

* enable tolerance and fields/particles optional comparison to fix tests

* more minor cleaning in space charge initialization tests

* update Python benchmarks for serial runs

* Run Python tests with 1 rank instead of 2

* fields that are zero should not be in reg tests

* Add developer documentation on checksum regression tests

Co-authored-by: Tools <warpx@lbl.gov>

1 files changed, 5 insertions, 0 deletions

diff --git a/Examples/Modules/RigidInjection/analysis_rigid_injection_LabFrame.py b/Examples/Modules/RigidInjection/analysis_rigid_injection_LabFrame.py
index 380212adb..2c97483fd 100755
--- a/Examples/Modules/RigidInjection/analysis_rigid_injection_LabFrame.py
+++ b/Examples/Modules/RigidInjection/analysis_rigid_injection_LabFrame.py
@@ -24,6 +24,8 @@ import sys
 import yt
 import numpy as np
 yt.funcs.mylog.setLevel(0)
+sys.path.insert(1, '../../../../warpx/Regression/Checksum/')
+import checksumAPI
 
 filename = sys.argv[1]
 
@@ -79,3 +81,6 @@ print("error_rel    : " + str(error_rel))
 print("tolerance_rel: " + str(tolerance_rel))
 
 assert( error_rel < tolerance_rel )
+
+test_name = filename[:-9] # Could also be os.path.split(os.getcwd())[1]
+checksumAPI.evaluate_checksum(test_name, filename)