HintCO/HintCOEngine/scripts/plot_results.py

import csv
import os
import sys
import matplotlib.pyplot as plt
from matplotlib.backends.backend_pdf import PdfPages
import argparse

def read_data(filepath):
	results = {}
	with open(filepath, 'r') as csvfile:
		reader = csv.DictReader(csvfile, delimiter=';')
		for col in reader.fieldnames:
			if col != "":
				results[col] = []
		for row in reader:
			for col in results:
				results[col].append(float(row[col]))
	return results

def selectTime(raw_data, maxTime):
    indexToTrim = 0
    maxIndex = len(raw_data["time"])
    while indexToTrim<maxIndex and raw_data["time"][indexToTrim] < maxTime:
        indexToTrim+=1
    
    if indexToTrim<maxIndex:
        for col in raw_data.keys():
            del raw_data[col][indexToTrim:]
        #print("Data trimmed up to index ", indexToTrim)
    
    return raw_data

def get_all_csvs(results_folder):
    results = []
    directory = os.fsencode(results_folder)
    for root, dirs, files in os.walk(directory):
        for file in files:
            filename = os.fsdecode(file)
            if filename.endswith(".csv"):
                filepath = os.fsdecode(os.path.join(root,file))
                results.append((filename, filepath))
    return results

def plot_results(results_folder, analyticalSolutionDir):
    for (filename, filepath) in get_all_csvs(results_folder):
        # Read all data onto a dictionary
        print("Reading data from file",filename,"... ", end="", flush=True)
        raw_data = read_data(filepath)
        print("Done.")
        
        if analyticalSolutionDir!="":
            analytical_sol_file = os.path.join(analyticalSolutionDir, filename)
            
            if os.path.exists(analytical_sol_file):                
                print("Reading analytical solution of file ",filename,"... ", end="", flush=True)
                analytical_data_raw = read_data(analytical_sol_file)
                print("Done.")
                
                print("Columns read: ", analytical_data_raw.keys())
                
                print("Trimming data to time ",raw_data["time"][-1],"... ", end="", flush=True)
                solution_raw = selectTime(analytical_data_raw, raw_data["time"][-1])
                print("Done.")
            else:
                print("No analytical solution for file ",filename,": " + analytical_sol_file)
                solution_raw = None
        else:
            solution_raw = None
         
            
        # Create a group of plots per filenames
        num_trajectories = len(raw_data.keys()) - 1 # Exclude time 
        print("Creating ",num_trajectories," plots... ", end="", flush=True)
        plot_num = 1
        plot_file_name = filepath.replace(".csv",".pdf")
        pp = PdfPages(plot_file_name)
        for col in raw_data.keys():
            if col != "time":
                print("Plotting col " + col + "...")
                plt.subplot(num_trajectories, 1, plot_num)
                plt.xlabel("time")
                plt.plot(raw_data["time"], raw_data[col], '-', label=col)
                if solution_raw != None:
                    if col in solution_raw:
                        plt.plot(solution_raw["time"], solution_raw[col], '-', label=col+"_anl")
                plt.legend()
                plot_num += 1
        pp.savefig()
        pp.close()
        plt.clf()
        
        print("Done.")

def smallest_identifier_string(strs):
    # Taken from https://medium.com/@d_dchris/10-methods-to-solve-the-longest-common-prefix-problem-using-python-leetcode-14-a87bb3eb0f3a
    longest_pre = ""
    if not strs: 
        return longest_pre
    shortest_str = min(strs, key=len)
    for i in range(len(shortest_str)):
        if all([x.startswith(shortest_str[:i+1]) for x in strs]):
            longest_pre = shortest_str[:i+1]
        else:
            break
    return longest_pre
  
def plot_merge_results(results_folder):
    # Get all files (and their path) ending in csv.
    all_csv_files = get_all_csvs(results_folder)
    
    # Group them by filename. Eg., all fmu1.csv files should be together.
    csv_files_grouped = {}
    for (filename, filepath) in all_csv_files:
        if not filename in csv_files_grouped:
            csv_files_grouped[filename] = []
        csv_files_grouped[filename].append(filepath)
    
    print("Files grouped.")
    
    # Create plot for each group
    for (filename, files) in csv_files_grouped.items():
        plot_filepath = files[0]
        plot_file_name = plot_filepath.replace(".csv",".pdf")
        pp = PdfPages(plot_file_name)
        group_id = smallest_identifier_string(files)
        
        for filepath in files:
            file_identifier = filepath[len(group_id):]
            raw_data = read_data(filepath)
            print("Read data from file",filename,".")
            
            num_trajectories = len(raw_data.keys()) - 1 # Exclude time 
            
            # Assumes that each csv column is in the same order on all csvs. This is a reasonable assumption...
            plot_num = 1
            for col in raw_data.keys():
                if col != "time":
                    print("Plotting col " + col + "...")
                    plt.subplot(num_trajectories, 1, plot_num)
                    plt.xlabel("time")
                    plt.plot(raw_data["time"], raw_data[col], '-', label=file_identifier+"."+col)
                    plt.legend()
                    plot_num += 1
        pp.savefig()
        pp.close()
        plt.clf()
    
    print("Done.")

if __name__ == '__main__':
    parser = argparse.ArgumentParser(description='HintCO plotting utility')
    parser.add_argument('resultsdir', help='directory where csv files of the co-simulation are stored.')
    parser.add_argument('--solutionsdir', help='directory where csv files of the analytical solution are stored.')
    parser.add_argument('--merge', action="store_true", help='Instead of creating a plot per folder, merge all plots into a single folder. This requires a fixed folder structure')
    
    args = parser.parse_args()
    
    results_folder = args.resultsdir
    analyticalSolutionDir = args.solutionsdir if args.solutionsdir else ""
    merge = args.merge
    print("results=",results_folder)
    print("analyticalSolutionDir=",analyticalSolutionDir)
    print("merge=",merge)
    if merge:
        plot_merge_results(results_folder)
    else:
        plot_results(results_folder, analyticalSolutionDir)