[PYTHON] Reading OpenFOAM time series data and sets data

Overview

I made a function to read OpenFOAM time series data and sets data (spatial distribution data output by the sample command) in Python, and simplified the plotting with matplotlib.

As a premise, the time series data of OpenFOAM is saved in a new directory every time it is restarted, but it is assumed that they are attached with the cat command to form a single file.

Python code

Function for reading time series data

• case is the case file name In + positions, you can put the location information of the time series data in a list.
• variables is a list of physical quantities (p, U, etc.) of time series data
• Read the time series data of each position at once.
• Returns dictionary data with the character string "physical quantity_location" as key and the actual data as ndarray as value.

def readHistory(case, positions, variables):
    import numpy as np
    
    """
    case: case name (string data)
    positions: position name (list data)
    variables: physical quantities (list data)
    """
    
    data ={}

    for position in positions:
        for variable in variables:
            key = variable+"_"+position
            value = np.genfromtxt(case+"/"+position+"/"+variable, unpack=True)
            data[key] = value
        
    return data

functions for reading sets data

Argument description

• case is the case file name
• file is the filename of the sets data A list of physical quantities (p, U, etc.) that + variables want to output
• sets Since data of multiple physical quantities is output to one file, each physical quantity is read individually.
• Returns dictionary data with the character string "physical quantity_location" as key and the actual data as dataframe as value.

def readSets(case, file, variables):
    import numpy as np
    import pandas as pd
    import glob
    import os
    
    """
    case: case name (string data)
    file: sets data file name (string data)
    variables: physical quantities (list data)
    """

    # list of timing
    list1 = glob.glob(case + "/sets/*")
    
    variable_names = file.split("_")
    # remove ".xy" 
    variable_names[-1] = variable_names[-1].split(".")[0]
    position, variable_names[0] = variable_names[0], "x"
    #print(variable_names)
        
    data = {}
    for variable in variables:
        key = variable+"_"+position
        data[key] = pd.DataFrame()
        
        for i in list1:
            path = i+"/" + file
            df = pd.read_table(path, header = None, names = variable_names)   
            data[key] = pd.concat([data[key], df[variable]], axis=1,)
    
        list_index=[]    
        for i in list1:
            list_index.append(str(float(os.path.basename(i))))
    
        data[key].columns = list_index
        data[key] = pd.concat([df[variable_names[0]], data[key].sort_index(axis=1)], axis=1)
    
    # return dictionary of dataframe
    return data

Plot with matplotlib

Plot the data with matplotlib.

Suppose the data layout is as follows. test01 is the case name.

test01/position-A/p test01/position-A/U test01/position-B/p test01/position-B/U test01/sets/0/center_p_k.xy test01/sets/10/center_p_k.xy

p and U are time series data of position-A and B center_p_k.xy is the distribution data at time 0 and 10. is.

There are multiple cases, and it is designed assuming that you can easily create a plot with only the case name changed.

import matplotlib.pyplot as plt
%matplotlib inline

case = "test01"
history_position = ["position-A", "position-B"]
history_variables = ["p", "U"]
sets_variables = ["p", "k"]
sets_position = "center"
dataSet_test01= [readHistory(case, history_position, history_variables), 
                  readSets(case, "center_p_k.xy", sets_variables)]

#Plot of a set of time series data
for i in history_position:
    for j in history_variables:
        fig = plt.figure(figsize=(8,8))
        ax = fig.add_subplot(111)

        ax.plot(dataSet_test01[0][j+"_"+i][0], dataSet_test01[0][j+"_"+i][1], label=j)
        ax.legend(bbox_to_anchor=(1.01, 1., 0., 0), loc='upper left', borderaxespad=0.,)

#Spatial distribution plot
for sets in sets_variables:
    for column in dataSet_test01[1][sets+"_"+sets_position]:
        fig = plt.figure(figsize=(8,8))
        ax = fig.add_subplot(111)
    
        ax.plot(dataSet_test01[1][sets+"_"+sets_position]["x"], 
            dataSet_test01[1][sets+"_"+sets_position][column], 
            label=sets+"_"+column)
        ax.legend(bbox_to_anchor=(1.01, 1., 0., 0), loc='upper left', borderaxespad=0.,)