Various minor improvements

Author: jorgensd

code/postprocess.py

@@ -1,12 +1,24 @@
-"""Script for converting file """
+"""
+Script for converting file
+
+SPDX-License-Identifier:    MIT
+"""
 import json
+import logging
+import sys
 import typing
 from pathlib import Path
 
 import config
 import dolfin
-from cardiac_geometries.geometry import Geometry
-from cardiac_geometries.geometry import load_microstructure
+import numpy as np
+from cardiac_geometries.geometry import Geometry, load_microstructure
+
+logger = logging.Logger(__name__, logging.INFO)
+ch = logging.StreamHandler(sys.stdout)
+FORMAT = '%(levelname)-5s [%(filename)s:%(lineno)d] %(message)s'
+ch.setFormatter(logging.Formatter(FORMAT))
+logger.addHandler(ch)
 
 
 def generate_fiber_xdmf_file(
@@ -17,17 +29,17 @@ def generate_fiber_xdmf_file(
 
     geo = Geometry.from_file(outfile, schema_path=outfile.with_suffix(".json"))
 
-    f0, s0, n0 = load_microstructure(
+    f0, _, _ = load_microstructure(
         mesh=geo.mesh,
         microstructure_path=microstructure_path,
     )
 
-    # Save fibers to a file that can be visualized in paraview
+    # Save fibers to a file that can be visualized in Paraview
 
     with dolfin.XDMFFile(fiber_path.as_posix()) as f:
         f.write(f0)
 
-    print(f"Saved fibers to {fiber_path}")
+    logger.info(f"Saved fibers to {fiber_path}")
     # Compute some features. This could be some results presented in the paper
 
     return {
@@ -39,11 +51,28 @@ def generate_fiber_xdmf_file(
     }
 
 
-def check_results(features_path: Path, features):
-    expected_feautres = json.loads(features_path.read_text())
-    print("Checking reproducibility")
-    assert expected_feautres == features
-    print("Results are reproducible!")
+def check_results(features_path: Path, features: dict):
+    """
+    Check if data from input file `features_path` matches input features
+
+    Args:
+        features_path (Path): Path to reference results
+        features (dict): Results generated from current simulation
+    """
+    expected_features = json.loads(features_path.read_text())
+    logger.info("Checking reproducibility")
+    reproducible = True
+    # Check each (key, value) pair in the features and check they are
+    # within machine precision
+    for key in expected_features.keys():
+        if not np.isclose(expected_features[key], features[key]):
+            logger.error(f"{key}: {expected_features[key]}!={features[key]}")
+            reproducible = False
+
+    if reproducible:
+        logger.info("Results are reproducible")
+    else:
+        raise RuntimeError("Results are not reproducible")
 
 
 def main() -> int:

code/pre_processing.py

@@ -16,11 +16,11 @@
 import sys
 
 # We set default log level to be info
-logger = logging.Logger("Preprocessing", level=logging.INFO)
-logger.addHandler(logging.StreamHandler(sys.stdout))
-# Mute FFC and UFL errors at this stage
-logging.getLogger('FFC').setLevel(logging.WARNING)
-logging.getLogger('UFL').setLevel(logging.WARNING)
+logger = logging.Logger(__name__, logging.INFO)
+ch = logging.StreamHandler(sys.stdout)
+FORMAT = '%(levelname)-5s [%(filename)s:%(lineno)d] %(message)s'
+ch.setFormatter(logging.Formatter(FORMAT))
+logger.addHandler(ch)
 
 schema = {
     "mesh": H5Path(

code/run_fiber_generation.py

@@ -3,14 +3,24 @@
 algorithm
 
 https://finsberg.github.io/ldrb/
+
+SPDX-License-Identifier:    MIT
 """
+import logging
+import sys
 from pathlib import Path
 
 import config
 import dolfin
 import ldrb
 from cardiac_geometries.geometry import Geometry
 
+logger = logging.Logger(__name__, logging.INFO)
+ch = logging.StreamHandler(sys.stdout)
+FORMAT = '%(levelname)-5s [%(filename)s:%(lineno)d] %(message)s'
+ch.setFormatter(logging.Formatter(FORMAT))
+logger.addHandler(ch)
+
 
 def generate_fibers(outfile: Path, microstructure_path: Path) -> None:
 
@@ -46,7 +56,7 @@ def generate_fibers(outfile: Path, microstructure_path: Path) -> None:
         h5file.write(s0, "s0")
         h5file.write(n0, "n0")
 
-    print(f"Microstructure saved to {microstructure_path}")
+    logger.info(f"Microstructure saved to {microstructure_path}")
 
 
 def main() -> int:

docs/_config.yml

@@ -35,6 +35,7 @@ parse:
 sphinx:
   config:
     bibtex_bibfiles: ["refs.bib"]
+    suppress_warnings: ["bibtex.duplicate_citation"] # If the same paper is cited in multiple files
 
   extra_extensions:
   - 'sphinx.ext.autodoc'

docs/abstract.md

@@ -5,7 +5,7 @@ Here you can write a little abstract for the paper or some text explaining what
 
 You could also add citation that you can add to the `refs.bib` in the `docs` folder. You can read more about adding citations [here](https://jupyterbook.org/en/stable/content/citations.html).
 
-In this example we use the LDRB algorithm {cite}`bayer2012novel` to data from {cite}`martinez2019repository`
+In this example we use the LDRB algorithm {cite}`bayer2012novel` to data from {cite}`martinez2019repository`.
 
 
 ## References

docs/reproducing.md

@@ -3,7 +3,7 @@
 
 ## Data
 
-Data is available in a dropbox folder. Use the script `download_data.sh` in the data folder to download the data, i.e
+Data is available in a Dropbox folder. Use the script `download_data.sh` in the data folder to download the data, i.e
 ```bash
 cd data
 bash download_data.sh
@@ -18,9 +18,9 @@ The data folder should have the following structure after the data is downloaded
     ├── heart01.msh
     └── heart02.msh
 ```
-These meshes are originally taken from <https://ora.ox.ac.uk/objects/uuid:951b086c-c4ba-41ef-b967-c2106d87ee06>, but since the original data is about 26GB we decided to make a smaller dataset for this example.
+These meshes are originally taken from {cite}`martinez2019repository`, but since the original data is about 26GB we decided to make a smaller dataset for this example.
 
-Eventually when you publish a paper you could put this data on e.g [Zenodo](https://zenodo.org). That will make sure the data gets it's own DOI.
+Eventually when you publish a paper you could put this data on e.g [Zenodo](https://zenodo.org). That will make sure the data gets it's own [Digital Object Identifier](https://www.doi.org/) (DOI).
 
 
 ## Scripts
@@ -31,7 +31,7 @@ In order to reproduce the results you need to first run the pre-processing scrip
 ```
 python3 pre_processing.py
 ```
-This will convert the meshes from Gmsh to a dolfin format.
+This will convert the meshes from Gmsh to a Dolfin format.
 
 ### Fiber generation
 The next step is to run the fiber generation. You can do this by running the script
@@ -46,3 +46,7 @@ The final step is to postprocess the results by running the script
 python3 postprocess.py
 ```
 This will generate a file for visualizing the fibers in the Paraview (inside `code/results` called  `fiber_<heart_nr>.xdmf`). This script will also compare some features computed from the fibers with the results published in the (artificial) paper. If the results differ, then the program will raise an error.
+
+```{bibliography}
+:filter: docname in docnames
+```