adding model benchmarking workflow

Author: valentina-s

.github/workflows/model_benchmarking.yml

@@ -0,0 +1,68 @@
+name: Model Benchmarking
+
+on:
+  pull request:
+    types: ['opened']
+
+env:
+  pr_username: github.event.pull_request.user.login
+  
+  # schedule:
+  #  - cron: '9 * * * *'
+    
+jobs:
+  process:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v4
+
+    - uses: actions/setup-python@v5
+      with:
+        python-version: '3.9.18'
+        cache: 'pip'
+        cache-dependency-path: ambient_sound_analysis/requirements_noise.txt
+
+    - name: Install Python dependencies
+      run: |
+        pip install -r ambient_sound_analysis/requirements_noise.txt
+
+ 
+    # - name: Install ffmpeg
+    #  run: |
+    #   sudo apt-get update
+    #   sudo apt-get install ffmpeg
+    
+    - uses: awalsh128/cache-apt-pkgs-action@latest
+      with:
+        packages: ffmpeg
+
+    - name: Create plots
+      if: steps.cache.outputs.cache-hit != 'true'
+      run: |
+        python ambient_sound_analysis/model_benchmarking.py
+
+
+
+    - name: Append username
+      run:|
+        sed 's/$/ $pr_username/' 'csv/2024-06-01 17:00:00.csv' > 'csv/2024-06-01 17:00:00_new.csv'
+
+    - name: Upload to GitHub repo
+      uses: stefanzweifel/git-auto-commit-action@v5
+      with:
+          commit_message: Commit to Github
+          file_pattern: 'ambient_sound_analysis/csv/*.csv'
+
+    #- name: Upload as artifact
+    #  uses: actions/upload-artifact@v4
+    # with:
+    #     name: ambient_sound_plots
+    #     path: ambient_sound_analysis/img/*.png
+
+    #- name: Upload to Google Drive
+    #  uses: AnimMouse/setup-rclone@v1
+    #  with:
+    #    rclone_config: ${{secrets.RCLONE_CONFIG}}
+      
+    # - run: |
+    #    rclone copy ambient_sound_analysis/img/broadband.png mydrive:rclone_uploads/

ambient_sound_analysis/model_benchmarking.py

@@ -0,0 +1,61 @@
+# importing general Python libraries
+import pandas as pd
+import datetime as dt
+import os
+import matplotlib.pyplot as plt
+import pytz
+import plotly.graph_objects as go
+import numpy as np
+
+# importing orcasound_noise libraries
+from orcasound_noise.pipeline.pipeline import NoiseAnalysisPipeline
+from orcasound_noise.utils import Hydrophone
+from orcasound_noise.pipeline.acoustic_util import plot_spec, plot_bb
+
+
+# Set Location and Resolution
+# Port Townsend, 1 Hz Frequency, 60-second samples
+if __name__ == '__main__':
+    pipeline = NoiseAnalysisPipeline(Hydrophone.ORCASOUND_LAB,
+                                     delta_f=10, bands=None,
+                                     delta_t=60, mode='safe')
+
+
+
+
+# Generate parquet dataframes with noise levels for a time period
+
+# now = dt.datetime.now(pytz.timezone('US/Pacific'))
+# fix time
+now = dt.datetime(2024, 6, 1, 17, 0, 0)
+psd_path, broadband_path = pipeline.generate_parquet_file(now - dt.timedelta(hours = 13), 
+                                                          now - dt.timedelta(hours = 8), 
+                                                          upload_to_s3=False)
+
+# Read the parquet files
+# psd_df = pd.read_parquet(psd_path)
+bb_df = pd.read_parquet(broadband_path)
+
+# set threshold
+threshold = 5
+
+import numpy as np
+nof_ships = (np.diff((bb_df['0']>threshold).astype('uint8'))==1).sum()
+
+
+# Create a new directory if it does not exist
+if not os.path.exists('ambient_sound_analysis/csv'):
+   os.makedirs('ambient_sound_analysis/csv')
+
+pd.DataFrame(nof_ships).to_csv('csv/'+str(now)+'.csv', header=False, index=False)
+
+
+
+
+# Create and save psd plot 
+# fig = plot_spec(psd_df)
+# fig.write_image('ambient_sound_analysis/img/psd.png')
+
+# Create and save bb plot
+# fig = plot_bb(bb_df)
+# fig.savefig('ambient_sound_analysis/img/broadband.png')