setup.py + cmd line script

MANIFEST.in

@@ -0,0 +1 @@
+recursive-include models *.h5 *.json

bin/crystalml

@@ -0,0 +1,13 @@
+#!/usr/bin/env pythonw
+import os
+import argparse
+
+from src.crystal_processing.process_image_folder import process_image_folder
+
+# Setup parser
+ap = argparse.ArgumentParser()
+ap.add_argument("directory", default=os.getcwd(), help="Path of the directory")
+
+args = ap.parse_args()
+
+process_image_folder(args.directory)

requirements.txt

@@ -5,9 +5,9 @@ opencv-python==4.0.0.21
 pandas==0.23.4
 Pillow==6.0.0
 plotly==3.6.1
+setuptools
 scikit-image==0.15.0
 scipy==1.1.0
 seaborn==0.9.0
 tensorboard==1.13.1
 tensorflow==1.13.1
-tqdm==4.28.1

setup.py

@@ -0,0 +1,35 @@
+from setuptools import setup, find_packages
+
+with open("README.md", "r") as f:
+    long_description = f.read()
+
+setup(name='crystalml',
+      version='0.0.1',
+      description='Integrated tool to measure the nucleation rate of protein crystals. ',
+      long_description=long_description,
+      long_description_content_type="text/markdown",
+      url='https://github.com/hlgirard/CrystalML',
+      author='Henri-Louis Girard',
+      author_email='[email protected]',
+      license='GPLv3',
+      packages=find_packages(exclude=['tests']),
+      install_requires=[
+          'tensorflow',
+          'matplotlib',
+          'numpy',
+          'opencv-python',
+          'pandas',
+          'pillow',
+          'plotly',
+          'scikit-image',
+          'setuptools',
+          'scipy',
+          'seaborn',
+          'tensorboard'
+
+      ],
+      scripts=[
+          'bin/crystalml',
+      ],
+      zip_safe=False,
+      include_package_data=True)

src/crystal_processing/process_image_folder.py

@@ -6,33 +6,37 @@
 import re
 import logging
 from joblib import Parallel, delayed
+import pkg_resources
+import logging
 
 import numpy as np
 import pandas as pd
 
 from tensorflow.keras.models import model_from_json
+logging.getLogger('tensorflow').disabled = True
 
 from src.data.utils import select_rectangle, get_date_taken, open_grey_scale_image
 from src.data.segment_droplets import crop, segment, extract_indiv_droplets
 from src.visualization.image_processing_overlay import save_overlay_image
 from src.visualization.process_plotting import plot_crystal_data
 
-def load_model(path):
+def load_model(model_name):
     '''Loads model from path and get most recent associated weights'''
     # TODO: Move this methods to the models utils package
 
-    model_name = path.split('/')[-1].split('.')[0]
+    model_basename = model_name.split('.')[0]
+    model_path = pkg_resources.resource_filename('models', model_name)
 
     ## Load model from JSON
-    with open(path, 'r') as json_file:
+    with open(model_path, 'r') as json_file:
         loaded_model_json = json_file.read()
 
     model = model_from_json(loaded_model_json)
 
     ## Load weights into model
-    model_list = sorted([model for model in os.listdir("models") if model.startswith(model_name) and model.endswith('.h5')], key = lambda x: int(re.search(r'\d+', x).group(0)))
+    model_list = sorted([model for model in pkg_resources.resource_listdir('models', '.') if model.startswith(model_basename) and model.endswith('.h5')], key = lambda x: int(re.search(r'\d+', x).group(0)))
     logging.info("Loading model weights: %s", model_list[-1])
-    model.load_weights("models/" + model_list[-1])
+    model.load_weights(pkg_resources.resource_filename('models', model_list[-1]))
 
     return model
 
@@ -92,7 +96,7 @@ def process_image(image_path, crop_box, model, save_overlay = False):
 
     return (date_taken, num_drops, num_clear, num_crystal)
 
-def process_image_batch(image_list, crop_box, model_path, save_overlay = False):
+def process_image_batch(image_list, crop_box, model_name, save_overlay = False):
     '''Process a batch of images and return a list of results
 
     Parameters
@@ -101,7 +105,7 @@ def process_image_batch(image_list, crop_box, model_path, save_overlay = False):
         List of paths to the image to process
     crop_box: (minRow, maxRow, minCol, maxCol)
         Cropping box to select the region of interest
-    model_path: string
+    model_name: string
         Path to the tensorflow model to load
     save_overlay: bool, optional
         Save an image with green / red overlays for drops containing crystals / empty to `image_path / overlay`
@@ -114,7 +118,7 @@ def process_image_batch(image_list, crop_box, model_path, save_overlay = False):
     '''
 
     # Instantiate the model
-    model = load_model(model_path)
+    model = load_model(model_name)
 
     # Process the data
     data = []
@@ -138,7 +142,7 @@ def process_image_folder(directory, crop_box=None, show_plot=False, save_overlay
     print(f"Number of batches: {num_batches}")
 
     # Define the model path
-    model_path = "models/cnn-simple-model.json"
+    model_name = "cnn-simple-model.json"
 
     # Obtain crop box from user if not passed as argument
     if not crop_box:
@@ -148,10 +152,10 @@ def process_image_folder(directory, crop_box=None, show_plot=False, save_overlay
     # Process all images from directory in parallel
     if num_batches == 0:
         # Process serialy
-        data = [process_image_batch(image_list[i*batch_size:min([(i+1)*batch_size, num_images])], crop_box, model_path, save_overlay)
+        data = [process_image_batch(image_list[i*batch_size:min([(i+1)*batch_size, num_images])], crop_box, model_name, save_overlay)
                 for i in range(num_batches)]
     else:
-        data = Parallel(n_jobs=-2, verbose=10)(delayed(process_image_batch)(image_list[i*batch_size:min([(i+1)*batch_size, num_images])], crop_box, model_path, save_overlay)
+        data = Parallel(n_jobs=-2, verbose=10)(delayed(process_image_batch)(image_list[i*batch_size:min([(i+1)*batch_size, num_images])], crop_box, model_name, save_overlay)
                                    for i in range(num_batches))
 
     flat_data = [item for sublist in data for item in sublist]
@@ -165,7 +169,6 @@ def process_image_folder(directory, crop_box=None, show_plot=False, save_overlay
     if show_plot:
         plot_crystal_data(df, directory)
 
-
 if __name__ == "__main__":
     folder = "notebooks/example_data"