1) Added tool to generate synthetic suitability data. 2) Added code t…

…o load and use that data. 3) Added code to run script to generate code if not found. 4) Added manifest.py file to be central store of paths to input and output datafiles.

MANIFEST.in

@@ -1,4 +1,5 @@
 include idmlaser_cholera/USA-pyramid-2023.csv
 include src/idmlaser_cholera/mods/libtx.so
 include src/idmlaser_cholera/mods/libages.so
+include src/idmlaser_cholera/tools/make_suitability_random_data.py 
 

src/idmlaser_cholera/cholera.py

@@ -4,6 +4,7 @@
 import numpy as np
 from datetime import datetime
 from tqdm import tqdm
+from . import manifest
 
 # Very simple!
 class Model:
@@ -119,11 +120,18 @@ def save_pops_in_nodes( model, nn_nodes, initial_populations):
 model.nodes.add_vector_property("network", model.nodes.count, dtype=np.float32)
 # The climatically driven environmental suitability of V. cholerae by node and time
 model.nodes.add_vector_property("psi", model.params.ticks, dtype=np.float32)
-model.nodes.psi[:] = 0.001 # placeholder, probably load from csv
+#model.nodes.psi[:] = 0.001 # placeholder, probably load from csv
 def init_psi_from_data():
-    suitability_filename = "suitability_data_random_419_5yr.csv"
+    suitability_filename = manifest.psi_data
     import pandas as pd 
-    data = pd.read_csv(suitability_filename)
+    try:
+        data = pd.read_csv(suitability_filename)
+    except Exception as ex:
+        print( str( ex ) )
+        print( "Running 'python -m idmlaser_cholera.tools.make_synthetic_suitability_data' and continuing." )
+        import idmlaser_cholera.tools.make_suitability_random_data
+        data = pd.read_csv(suitability_filename)
+
 
     # Convert the DataFrame into a NumPy array
     suitability_data = data.values
@@ -133,7 +141,7 @@ def init_psi_from_data():
 
     # Assign the data to the "psi" vector in the model
     model.nodes.psi[:] = suitability_data
-#init_psi_from_data()
+init_psi_from_data()
 # theta: The proportion of the population that have adequate Water, Sanitation and Hygiene (WASH).
 model.nodes.add_scalar_property("WASH_fraction", dtype=np.float32) # leave at 0 for now, not used yet
 

src/idmlaser_cholera/manifest.py

@@ -0,0 +1 @@
+psi_data="synthetic_psi_data.csv"

src/idmlaser_cholera/tools/make_suitability_random_data.py

@@ -0,0 +1,34 @@
+import numpy as np
+import pandas as pd
+
+# Define the dimensions of the CSV
+rows = 419  # Number of nodes
+columns = 365 * 5  # Time steps (5 years with 365 steps per year)
+
+# Generate random amplitude (between 0.1 and 0.9) for each row
+amplitudes = np.random.uniform(0.1, 0.9, size=(rows, 1))
+
+# Generate random phase offsets (between 0 and 2*pi for a full cycle) for each row
+phase_offsets = np.random.uniform(0, 2 * np.pi, size=(rows, 1))
+
+# Create a time vector representing the time steps for the sin wave (columns represent days)
+time_vector = np.arange(columns)
+
+# Generate sinusoidal data for each row
+# Formula: value = amplitude * sin(2 * pi * time / period + phase) + 0.5 (to shift range to 0 to 1)
+sin_data = amplitudes * np.sin(2 * np.pi * time_vector / 365 + phase_offsets) + 0.5
+
+# Clip the data to ensure it stays between 0 and 1
+sin_data_clipped = np.clip(sin_data, 0, 1)
+
+# Create a DataFrame from the sinusoidal values
+df = pd.DataFrame(sin_data_clipped)
+
+# Define the filename for the CSV
+csv_filename = "synthetic_psi_data.csv"
+
+# Save the DataFrame to a CSV file
+df.to_csv(csv_filename, index=False)
+
+csv_filename
+