move some regression tests and utils into the repo

test/regression_tests/.gitignore

@@ -0,0 +1,2 @@
+*.log
+results.exo

test/regression_tests/run_regression_tests.py

@@ -0,0 +1,295 @@
+#!/usr/bin/env python3
+
+import argparse
+import glob
+import os
+import sys
+import time
+
+import yaml
+
+# Add the parent directory to the system path to import custom modules
+sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__), "..")))
+
+from utils.regression_test import ExodiffCheck, PeakMemoryCheck, RegressionTest
+
+
+def get_inputs_from_yaml_node(yaml_node, test_name_prefix, build_dir):
+    """
+    Extracts test inputs from a YAML node.
+    """
+    inputs = {
+        "test_name": f"{test_name_prefix}_{yaml_node['hardware']}_np_{yaml_node['num_processors']}",
+        "input_file": yaml_node["input_file"],
+        "peak_memory": yaml_node.get("peak_memory_check", {}).get("value"),
+        "peak_memory_percent_tolerance": yaml_node.get("peak_memory_check", {}).get(
+            "percent_tolerance"
+        ),
+        "exodiff": [
+            {
+                "compare_file": exodiff["compare_file"],
+                "results_file": exodiff["results_file"],
+                "gold_file": exodiff["gold_file"],
+            }
+            for exodiff in yaml_node["exodiff"]
+        ],
+        "executable_path": os.path.join(
+            build_dir,
+            "Release_gpu" if yaml_node["hardware"] == "gpu" else "Release",
+            "aperi-mech",
+        ),
+        "num_processors": yaml_node["num_processors"],
+    }
+    return inputs
+
+
+def expand_wildcards(file_patterns):
+    """
+    Expands wildcard patterns to a list of matching files.
+    """
+    expanded_files = []
+    for pattern in file_patterns:
+        expanded_files.extend(glob.glob(pattern))
+    return expanded_files
+
+
+def should_run_test(test_config, cpu_only, serial_only, parallel_only, gpu_only):
+    """
+    Determines if a test should be run based on the provided flags.
+    """
+    if (
+        (cpu_only and test_config["hardware"] != "cpu")
+        or (serial_only and test_config["num_processors"] != 1)
+        or (parallel_only and test_config["num_processors"] == 1)
+        or (gpu_only and test_config["hardware"] != "gpu")
+    ):
+        return False
+    return True
+
+
+def execute_test(test_config, dirpath, build_dir, keep_results):
+    """
+    Executes a single test and returns whether it passed.
+    """
+    print(f"  Running test {test_config['hardware']}_{test_config['num_processors']}")
+    inputs = get_inputs_from_yaml_node(
+        test_config, os.path.basename(dirpath), build_dir
+    )
+    regression_test = RegressionTest(
+        inputs["test_name"],
+        inputs["executable_path"],
+        inputs["num_processors"],
+        [inputs["input_file"]],
+    )
+    return_code, stats = regression_test.run()
+
+    if return_code != 0:
+        print("\033[91m  FAIL\033[0m")
+        return False
+
+    all_exodiff_passed = all(
+        ExodiffCheck(
+            f"{inputs['test_name']}_exodiff_{i}",
+            "exodiff",
+            exodiff["compare_file"],
+            exodiff["results_file"],
+            exodiff["gold_file"],
+            [],
+        ).run()
+        == 0
+        for i, exodiff in enumerate(inputs["exodiff"])
+    )
+
+    memcheck_passed = True
+    if inputs["peak_memory"] is not None:
+        peak_memory_check = PeakMemoryCheck(
+            f"{inputs['test_name']}_peak_memory",
+            stats["peak_memory"],
+            inputs["peak_memory"],
+            inputs["peak_memory_percent_tolerance"],
+        )
+        memcheck_passed = peak_memory_check.run() == 0
+
+    if all_exodiff_passed and memcheck_passed:
+        print("\033[92m  PASS\033[0m")
+        if (
+            not keep_results
+            and "cleanup" in test_config
+            and "remove" in test_config["cleanup"]
+        ):
+            cleanup_test_results(test_config)
+        return True
+    else:
+        print("\033[91m  FAIL\033[0m")
+        return False
+
+
+def cleanup_test_results(test_config):
+    """
+    Cleans up test results based on the cleanup configuration.
+    """
+    for item in test_config["cleanup"]["remove"]:
+        for file in expand_wildcards([item]):
+            os.remove(file)
+
+
+def run_regression_tests_from_directory(
+    root_dir,
+    build_dir,
+    cpu_only=False,
+    serial_only=False,
+    parallel_only=False,
+    gpu_only=False,
+    keep_results=False,
+):
+    """
+    Runs regression tests from the specified directory.
+    """
+    passing_tests = 0
+    total_tests = 0
+    current_dir = os.getcwd()
+
+    for dirpath, _, filenames in os.walk(root_dir):
+        if "test.yaml" in filenames:
+            os.chdir(dirpath)
+            print("-----------------------------------")
+            print(f"Running tests in {dirpath}")
+
+            with open("test.yaml", "r") as file:
+                yaml_node = yaml.safe_load(file)
+                test_configs = yaml_node["tests"]
+
+                for test_config in test_configs:
+                    if not should_run_test(
+                        test_config, cpu_only, serial_only, parallel_only, gpu_only
+                    ):
+                        continue
+
+                    if execute_test(test_config, dirpath, build_dir, keep_results):
+                        passing_tests += 1
+                    total_tests += 1
+
+            print("-----------------------------------\n")
+            os.chdir(current_dir)
+
+    return passing_tests, total_tests
+
+
+def clean_logs(root_dir):
+    """
+    Cleans log files from the specified directory.
+    """
+    for dirpath, _, filenames in os.walk(root_dir):
+        if "test.yaml" in filenames:
+            print("-----------------------------------")
+            print(f"Cleaning logs in {dirpath}")
+            for log_file in glob.glob(f"{dirpath}/*.log"):
+                os.remove(log_file)
+            print("-----------------------------------\n")
+
+
+def clean_results(root_dir):
+    """
+    Cleans result files from the specified directory.
+    """
+    for dirpath, _, filenames in os.walk(root_dir):
+        if "test.yaml" in filenames:
+            print("-----------------------------------")
+            print(f"Cleaning results in {dirpath}")
+            yaml_file = os.path.join(dirpath, "test.yaml")
+            with open(yaml_file, "r") as file:
+                yaml_node = yaml.safe_load(file)
+                test_configs = yaml_node["tests"]
+                for test_config in test_configs:
+                    if "cleanup" in test_config and "remove" in test_config["cleanup"]:
+                        for item in test_config["cleanup"]["remove"]:
+                            for file in expand_wildcards([item]):
+                                os.remove(os.path.join(dirpath, file))
+            print("-----------------------------------\n")
+
+
+def parse_arguments():
+    """
+    Parses command-line arguments.
+    """
+    parser = argparse.ArgumentParser(description="Run regression tests.")
+    parser.add_argument(
+        "--directory",
+        help="Directory root containing the tests. Will recursively search for test.yaml files.",
+        default=".",
+    )
+    parser.add_argument(
+        "--build_dir",
+        help="Directory containing the build",
+        default="/home/azureuser/projects/aperi-mech/build/",
+    )
+    parser.add_argument(
+        "--clean_logs",
+        help="Clean the log files from the tests and exit",
+        action="store_true",
+    )
+    parser.add_argument(
+        "--clean_results",
+        help="Clean the results files from the tests and exit",
+        action="store_true",
+    )
+    parser.add_argument(
+        "--keep_results",
+        help="Keep the results files even if the test passes. (results files are always kept if the test fails)",
+        action="store_true",
+    )
+    parser.add_argument(
+        "--no_preclean",
+        help="Do not clean the logs and results before running the tests",
+        action="store_true",
+    )
+    parser.add_argument("--cpu", help="Only run CPU tests", action="store_true")
+    parser.add_argument("--serial", help="Only run serial tests", action="store_true")
+    parser.add_argument(
+        "--parallel", help="Only run parallel tests", action="store_true"
+    )
+    parser.add_argument("--gpu", help="Only run GPU tests", action="store_true")
+    return parser.parse_args()
+
+
+if __name__ == "__main__":
+    args = parse_arguments()
+    build_dir = os.path.abspath(args.build_dir)
+    directory = os.path.abspath(args.directory)
+
+    # Clean logs and/or results if specified
+    if args.clean_logs or args.clean_results:
+        if args.clean_logs:
+            clean_logs(directory)
+        if args.clean_results:
+            clean_results(directory)
+        sys.exit(0)
+
+    # Pre-clean logs and results unless specified otherwise
+    if not args.no_preclean:
+        print("Cleaning logs and results before running the tests")
+        clean_logs(directory)
+        clean_results(directory)
+
+    # Run regression tests and measure time
+    start_time = time.perf_counter()
+    passing_tests, total_tests = run_regression_tests_from_directory(
+        directory,
+        build_dir,
+        args.cpu,
+        args.serial,
+        args.parallel,
+        args.gpu,
+        args.keep_results,
+    )
+    end_time = time.perf_counter()
+    print(f"Total time: {end_time - start_time:.4e} seconds")
+
+    failing_tests = total_tests - passing_tests
+    if failing_tests > 0:
+        print(f"{failing_tests} tests failed.")
+        print(f"{passing_tests} tests passed.")
+        sys.exit(1)
+    else:
+        print(f"All {passing_tests} tests passed.")
+        sys.exit(0)

test/regression_tests/taylor_bar/fem/compare.exodiff

@@ -0,0 +1,57 @@
+
+#  *****************************************************************
+#             EXODIFF	(Version: 3.33) Modified: 2024-03-13
+#             Authors:  Richard Drake, [email protected]           
+#                       Greg Sjaardema, [email protected]          
+#             Run on    2024/07/13   00:07:11 UTC
+#  *****************************************************************
+
+#  FILE 1: /home/azureuser/projects/aperi-mech_test/cylindrical_taylor_bar/regression/fem/gold_results.exo
+#   Title: IOSS Default Output Title
+#          Dim = 3, Nodes = 285, Elements = 904, Faces = 0, Edges = 0
+#          Element Blocks = 1, Face Blocks = 0, Edge Blocks = 0, Nodesets = 1, Sidesets = 0, Assemblies = 0
+#    Vars: Global = 0, Nodal = 18, Element = 1, Face = 0, Edge = 0, Nodeset = 0, Sideset = 0, Times = 2
+
+
+# ==============================================================
+#  NOTE: All node and element ids are reported as global ids.
+
+# NOTES:  - The min/max values are reporting the min/max in absolute value.
+#         - Time values (t) are 1-offset time step numbers.
+#         - Element block numbers are the block ids.
+#         - Node(n) and element(e) numbers are 1-offset.
+
+COORDINATES absolute 1.e-6    # min separation not calculated
+
+TIME STEPS relative 1.e-6 floor 0.0     # min:               0 @ t1 max:          0.0001 @ t2
+
+# No GLOBAL VARIABLES
+
+NODAL VARIABLES relative 1.e-6 floor 0.0
+	displacement_x         # min:               0 @ t1,n1	max:    0.0037303132 @ t2,n16
+	displacement_y         # min:               0 @ t1,n1	max:    0.0034884944 @ t2,n127
+	displacement_z         # min:               0 @ t1,n1	max:     0.037275041 @ t2,n210
+	acceleration_x         # min:               0 @ t1,n1	max:        21336627 @ t2,n229
+	acceleration_y         # min:               0 @ t1,n1	max:        33554769 @ t2,n29
+	acceleration_z         # min:               0 @ t1,n1	max:        10711345 @ t2,n48
+	force_x                # min:               0 @ t1,n1	max:       226751.51 @ t2,n276
+	force_y                # min:               0 @ t1,n1	max:       233764.41 @ t2,n259
+	force_z                # min:               0 @ t1,n1	max:       351701.62 @ t2,n233
+	mass_x                 # min:     0.001125266 @ t1,n3	max:     0.041543172 @ t1,n259
+	mass_y                 # min:     0.001125266 @ t1,n3	max:     0.041543172 @ t1,n259
+	mass_z                 # min:     0.001125266 @ t1,n3	max:     0.041543172 @ t1,n259
+	mass_from_elements_x   # min:     0.001125266 @ t1,n3	max:     0.041543172 @ t1,n259
+	mass_from_elements_y   # min:     0.001125266 @ t1,n3	max:     0.041543172 @ t1,n259
+	mass_from_elements_z   # min:     0.001125266 @ t1,n3	max:     0.041543172 @ t1,n259
+	velocity_x             # min:               0 @ t1,n1	max:       96.091629 @ t2,n200
+	velocity_y             # min:               0 @ t1,n1	max:       123.97411 @ t2,n32
+	velocity_z             # min:               0 @ t2,n2	max:             373 @ t1,n1
+
+# No NODESET VARIABLES
+
+# No SIDESET VARIABLES
+
+# No EDGE BLOCK VARIABLES
+
+# No FACE BLOCK VARIABLES
+