In [1]:

    

import pandas as pd
import numpy as np

# visualization libraries
import matplotlib.pyplot as plt
from matplotlib import gridspec

# plot the visuals in ipython
%matplotlib inline


    

In [9]:

    

input_dir = "/Users/tony/pypge/data/benchmarks/explicit/"
output_dir = "/Users/tony/pypge/experiments/output/basic/explicit/"

problem_name = "Koza_01"
prob_out_dir = output_dir + "clean/" + problem_name + "/out/F_linear__I_low__G_low/"

input_file = input_dir + problem_name + "_clean.csv"
in_df = pd.read_csv(input_file, skipinitialspace=True)

err_file = prob_out_dir + "pge_errs.log"
err_df = pd.read_csv(err_file, delim_whitespace=True)

final_file = prob_out_dir + "pge_final.log"
final_df = pd.read_csv(final_file, skipinitialspace=True)
print "loaded"

print in_df.columns
print err_df.columns
print final_df.columns


    

    




loaded
Index([u'x', u'out'], dtype='object')
Index([u'iteration', u'peekd_models', u'evald_models', u'peek_fit_loops',
       u'peek_point_evals', u'eval_fit_loops', u'eval_point_evals',
       u'total_point_evals', u'best_size', u'best_err', u'best_r2',
       u'best_vari', u'ave_size', u'ave_err', u'ave_r2', u'ave_evar'],
      dtype='object')
Index([u'id', u'sz', u'error', u'r2', u'expld_vari', u'theModel'], dtype='object')

In [23]:

    

final_df.plot(kind="scatter", x="sz", y="error")
print "done"


    

    




done






    

In [25]:

    

print err_df.columns
err_df['best_err'].plot(logy=True)
plt.show()


    

    




Index([u'iteration', u'peekd_models', u'evald_models', u'peek_fit_loops',
       u'peek_point_evals', u'eval_fit_loops', u'eval_point_evals',
       u'total_point_evals', u'best_size', u'best_err', u'best_r2',
       u'best_vari', u'ave_size', u'ave_err', u'ave_r2', u'ave_evar'],
      dtype='object')