Initial commit of minimum viable product (can generate a run and pull output tables as CSV/Excel format) for Python implementation.

Author: scottyunho

.gitignore

@@ -0,0 +1,6 @@
+build/
+dist/
+.vs/
+/PohemX-python.spec
+Model_Outputs/
+Functions/__pycache__

Functions/Create.py

@@ -0,0 +1,86 @@
+#To-do: Finished logic but adjust code to work as a module
+"""
+Author: Scott Yun Ho
+Date: Oct. 29, 2024
+Description: Functions that create something new on the server-side.
+"""
+
+import requests
+import pandas as pd
+import time
+
+
+"""
+Description: Create and run a given model on OpenM++ engine. Send a request to run model, then once finished, return model_run_status.
+Parameters: 
+Return: 
+"""
+
+def create_model_run(model_digest, oms_url, simulation_cases=1e5, tables=[], threads=1, sub_samples=1) :
+
+    # Coerce numeric parameters to strings
+    simulation_cases = str(int(simulation_cases))
+    threads = str(int(threads))
+    sub_samples = str(int(sub_samples))
+
+    # Set JSON request
+    json_request = {
+        'ModelName': model_digest,
+        'Opts': {
+            'Parameter.SimulationCases': simulation_cases,
+            'OpenM.Threads': threads,
+            'OpenM.SubValues': sub_samples,
+            'OpenM.LogToConsole': 'true',
+            'OpenM.ProgressPercent': '100'
+        },
+        'Tables': tables
+    }
+
+    # Send request to oms
+    print("Sending API request to web service...")
+    x = requests.post(oms_url + '/api/run', json = json_request)
+
+    # Set model run status
+    model_run_status = ''
+
+    startTime = time.time() #have start time, then have currentTime in model progress to repeatedly check how much time has elapsed.
+    # Monitor model run status
+    while model_run_status in '' 'i' 'p' 'w' :
+      
+        # Pause for one second to give time for information to be delivered. 
+        time.sleep(1)
+    
+        # Get current model run status
+        model_run_status = requests.get(oms_url + '/api/model/' + model_digest + '/run/status/last').json()['Status']
+        currentTime = time.time()
+
+        # Render conditional message to console 
+        if model_run_status in '' 'i' :
+            print("Waiting for model run to start. ", end="")
+            print( "Current time elapsed:", round((currentTime - startTime), 2), "seconds so far... ")
+            continue
+
+        elif model_run_status == 's' :
+            print("Model run completed successfully! ", end="")
+            print("Final time elapsed:", round((currentTime - startTime), 2), "seconds. \n")
+            break
+
+        elif model_run_status in 'i' 'p' 'w' :
+            print("Model run in progress. ", end="")
+            print("Current time elapsed:", round((currentTime - startTime), 2), "seconds so far... ")
+            continue
+
+        elif model_run_status == 'e' :
+            print("Model run completed, but with errors. :(")
+            print("Final time elapsed:", round((currentTime - startTime), 2), "seconds. \n")
+            break
+
+    return model_run_status
+
+"""
+Description: 
+Parameters: 
+Return: 
+"""
+def create_scenario():
+    return 1

Functions/Delete.py

@@ -0,0 +1,22 @@
+"""
+Author: Scott Yun Ho
+Date: Oct. 29, 2024
+Description: Functions that delete something existing on the server-side.
+"""
+
+
+"""
+Description: 
+Parameters: 
+Return: 
+"""
+def delete_model_run():
+	return 1
+
+"""
+Description: 
+Parameters: 
+Return: 
+"""
+def delete_scenario():
+	return 1

Functions/Get.py

@@ -0,0 +1,36 @@
+"""
+Author: Scott Yun Ho
+Date: Oct. 29, 2024
+Description: Functions that get unspecific information from OpenM++.
+"""
+
+import requests
+import pandas as pd
+
+
+"""
+Description: Provided that the OpenM++ web service (the "oms") is running at oms_url, get list of models from the API route oms_url/api/model-list.
+Parameters: 
+Return: 
+"""
+def get_models(oms_url) :
+    x = requests.get(oms_url + '/api/model-list').json()
+    return pd.DataFrame([item['Model'] for item in x])
+
+
+"""
+Description: Provided that the OpenM++ web service (the "oms") is running at oms_url, get full list of model runs from API route at oms_url/api/model/model_digest/run-list.
+Parameters: 
+Return: 
+"""
+def get_model_runs(oms_url, model_digest):
+    x = requests.get(oms_url + '/api/model/' + model_digest + '/run-list').json()
+    return pd.DataFrame([item for item in x])
+
+"""
+Description: 
+Parameters: 
+Return: 
+"""
+def get_scenarios():
+    return 1

Functions/Load.py

@@ -0,0 +1,119 @@
+"""
+Author: Scott Yun Ho
+Date: Oct. 29, 2024
+Description: Functions that get specific information from OpenM++.
+"""
+import Functions.Get as Get
+
+import requests
+import pandas as pd
+import csv
+
+"""
+Description: Retrieve specific model run based on model_name and model_digest. Requires model_run_status from Create_model_run.py.
+Parameters: 
+Return: 
+"""
+def load_model_run(model_name, model_digest, model_run_status, oms_url, tables):
+
+    # Create model run object. Run has now been completed, so create an object that gathers the finished information in one place. 
+    model_run = {
+        'model': model_name, 
+        'digest': Get.get_model_runs(oms_url, model_digest).iloc[-1]['RunDigest'], 
+        'status': model_run_status
+    }
+
+    # Retrieve specified output tables after model run.
+    print(tables)
+    for table in tables:
+        print('Retrieving ' + table + ' table...')
+        model_run[table] = load_output_table(model_digest=model_digest, model_run_digest=model_run['digest'], table=table, oms_url=oms_url)
+
+    return model_run
+
+#To-do: finished logic here but test
+"""
+Description: 
+Parameters: 
+Return: 
+"""
+def load_model(model_name):
+    x = requests.get(oms_url + '/api/model-list').json()
+    model_list = pd.DataFrame([item['Model'] for item in x])
+    return model_list.loc[model_list['ModelName'] == model_name]
+
+#To-do: get multiple specific runs. Take array as input and output info for multiple model runs.
+"""
+Description: 
+Parameters: 
+Return: 
+"""
+def load_model_runs():
+    return 1
+
+"""
+Description: 
+Parameters: 
+Return: 
+"""
+def load_output_table(model_digest, model_run_digest, table, oms_url):
+    # Get list from CSV format (data is neater). URL returns CSV file attachment in a response stream, which is why stream is necessary below. Informed by: https://stackoverflow.com/questions/35371043/use-python-requests-to-download-csv#:~:text=CSV_URL%20%3D%20%27http%3A//samplecsvs.s3.amazonaws.com/Sacramentorealestatetransactions.csv%27-,with%20requests,-.Session()%20as%20s%3A%0A%20%20%20%20download%20%3D%20s.get(CSV_URL)%0A%0A%20%20%20%20decoded_content
+    print("Getting response stream output and processing into a dataframe...")
+    raw_csv_data = []
+    with requests.Session() as stream:
+        download = stream.get(oms_url + '/api/model/' + model_digest + '/run/' + model_run_digest + '/table/' + table + '/expr/csv')
+        decoded_content = download.content.decode('utf-8')
+        cr = csv.reader(decoded_content.splitlines(), delimiter=',')
+        raw_csv_data = list(cr)
+
+    # Convert list of lists table output (raw_csv_data) to pandas dataframe.
+    headers = raw_csv_data.pop(0)
+    long_format_df = pd.DataFrame(raw_csv_data, columns=headers)
+
+    # Clean data to remove negative and positive infinity. regex=True needed to recognize the unicode for infinity: \u221e
+    #print("Cleaning infinity symbols from data...")
+    #long_format_df = long_format_df.replace("-"+"\u221e", "MIN", regex=True) 
+    #long_format_df = long_format_df.replace("\u221e", "MAX", regex=True)
+
+    # Previous format that returned everything in dims col.
+    """
+    #response = pd.DataFrame(requests.get(oms_url + '/api/model/' + model_digest + '/run/' + model_run_digest + '/table/' + table + '/expr').json())
+    #print(pd.DataFrame(requests.get(oms_url + '/api/model/' + model_digest + '/run/' + model_run_digest + '/table/' + table + '/expr/csv')))
+    #response = requests.get(oms_url + '/api/model/' + model_digest + '/run/' + model_run_digest + '/table/' + table + '/expr/csv')
+    #print(response.headers)
+    #output_df = pd.DataFrame(response.content)
+    #print(response)
+    #output_df = response
+
+    #output_df["Dims"] = output_df["Dims"].astype(str)  # Program cannot pivot using a list as index, so Dims col must be converted to string first.
+    #output_df = output_df.pivot(index="Dims", columns="ExprId", values="Value")
+    #print(output_df)
+    """
+
+    # Copy headers to use as columns during pivot, but not including expr_name and expr_value since we're pivoting on these values.
+    wide_format_indices = headers
+    wide_format_indices.remove("expr_name")
+    wide_format_indices.remove("expr_value")
+
+    # Pivot dataframe to be wide format.
+    print("Converting data to wide format...\n")
+    wide_format_df = long_format_df
+    wide_format_df = pd.pivot(wide_format_df, 
+        index=wide_format_indices, 
+        columns="expr_name", 
+        values="expr_value"
+    )
+
+    return wide_format_df
+
+"""
+Description: 
+Parameters: 
+Return: 
+"""
+def load_scenario():
+    return 1
+
+
+def load_model_digest(df, model_name) :
+    return df['Digest'][df.loc[df['Name'] == model_name].index.values[0]]

Functions/Utilities.py

@@ -0,0 +1,58 @@
+"""
+Author: Scott Yun Ho
+Date: Oct. 29, 2024
+Description: Helper functions that don't directly perform an API call or don't interact with OpenM++ in general.
+"""
+
+import pandas as pd
+import os
+
+
+
+"""
+Description: 
+Parameters: 
+Return: 
+"""
+def download_csv(file_name, df_table):
+    return df_table.to_csv(file_name, index=True)
+
+"""
+Description: 
+Parameters: 
+Return: 
+"""
+def download_excel(file_name, df_table):
+    return df_table.to_excel(file_name)
+
+# Create folder within current directory and switch to directory for output download.
+"""
+Description: 
+Parameters: 
+Return: 
+"""
+def create_output_folder(output_folder_name):
+    current_dir = os.getcwd()
+    folder_destination = os.path.join(current_dir, output_folder_name)
+    if not os.path.exists(folder_destination):
+        os.makedirs(folder_destination)
+    os.chdir(output_folder_name)
+
+"""
+Description: 
+Parameters: 
+Return: 
+"""
+def generate_output_files(model_run, tables, output_folder_name):
+    # Create output folder.
+    create_output_folder(output_folder_name)
+
+    # Get and download output tables specified in tables list.
+    for table in tables:
+        output_table = model_run[table]
+        #print(table)
+        print("Downloading " + table + " in CSV format.")
+        download_csv(table + ".csv", output_table)
+        print("Downloading " + table + " in Excel format.")
+        download_excel(table + ".xlsx", output_table)
+    print("\n")