#!/usr/bin/env python import matplotlib.cm as cmx import matplotlib.colors as colors import matplotlib.pyplot as plt import numpy as np import pandas as pd import optparse import re import yaml def nonlinear_history_iterations(yaml_data, mode, ax = None): """Show number of linear iterations per nonlinear step across the whole simulation""" ticks = [] itit = [] maxits = 0 offset = 0 mx = 0 i = 0 for step in sorted(yaml_data['Steps']): c_step = yaml_data['Steps'][step] # check whether nl_its is contained in list # if YES, it is a dynamic simulation, i.e. Steps contains time step (transient steps) # if NO, it is a static simulation with nonlinear iterations only bDynamicSim = True try: c_step['nl_its'] except KeyError: bDynamicSim = False if bDynamicSim == False: raise RuntimeError("Static simulations not supported, yet") its = [] for nlstep in sorted(c_step): if nlstep == 'nl_its': continue ticks.append(str(len(its))) its.append(c_step[nlstep]['its']) if maxits < len(its): maxits = len(its) itit.append(its) i += 1 print itit ikk = [] for k in range(maxits): ik = [] for l in range(len(itit)): its = itit[l] if len(its) > k: ik.append(its[k]) else: ik.append(0) ikk.append(ik) ind = np.arange(len(itit)) bottomvals = [] for l in range(len(itit)): bottomvals.append(0) for k in range(maxits): p1 = plt.bar(ind, ikk[k], 1.0, color='r',bottom = bottomvals) for l in range(len(itit)): bottomvals[l] = bottomvals[l] + ikk[k][l] ax.set_ylabel('Number of accumulated linear iterations') ax.set_xlabel('Nonlinear iteration index') plt.show() if __name__ == '__main__': p = optparse.OptionParser() # action arguments p.add_option('-i', '--input-file', dest='input_file') p.add_option('-o', '--output-file', dest='output_file') p.add_option('-d', '--display', dest='display', default='print') p.add_option('-a', '--analysis', dest='analysis', default='mode1') # parse options, arguments = p.parse_args() # validate options if options.input_file == None: raise RuntimeError("Please specify the input file") with open(options.input_file) as data_file: yaml_data = yaml.safe_load(data_file) f, ax = plt.subplots(1) analysis = options.analysis display = options.display nonlinear_history_iterations(yaml_data, mode=display, ax=ax) if options.output_file != None: plt.savefig(options.output_file, bbox_inches='tight') elif display == 'display': plt.show()