Application of Python Alternating Conditional Expectation (ACE) algorithm

Question

I am trying to apply the ACE algorithm outlined here: https://pypi.python.org/pypi/ace/0.3

Here is the code from the sample page, which works:

from ace.samples import wang04
x, y = wang04.build_sample_ace_problem_wang04(N=200)

from ace import model
myace = model.Model()
myace.build_model_from_xy(x, y)

Here is the code that I am trying to run:

import pandas as pd
import numpy as np
from ace import model

filename = 'G_perm.txt'

well = pd.read_csv(filename, sep="\s+", error_bad_lines=False)
well.columns = ["k", "por"]

x = np.zeros(len(well))
y = np.zeros(len(well))

for i in range(len(well)):
    x[i] = well.k[i] 
    y[i] = well.por[i]

myace = model.Model()
myace.build_model_from_xy(x, y)

Here is a sample of the txt file, with 162 lines:

k       por
306.0   26.3
61.0    25.4
1059.0  30.9
1120.0  30.0
540.0   29.8
272.0   27.0
430.0   28.3
84.0    25.6
788.0   30.2
490.0   28.7
541.0   28.8

When I try running the code, I get list index out of range Has anyone implemented ACE successfully with Pandas and Numpy?

Answer 1

The x_values argument for build_model_from_xy appears to require an iterator, as if it is iterating through the list of x variables. So, the following replacement for your last line should work:

myace.build_model_from_xy([x], y)

Note that the x from the sample code (wang04.build_sample_ace_problem_wang04(N=200)) is, in fact, a list of arrays.