import numpy as np
from repository import Repository
from configuration import config
from sklearn.cross_validation import train_test_split
from sklearn.cross_validation import ShuffleSplit
from sklearn import tree
from sklearn.decomposition import PCA
from geopy.distance import vincenty
import matplotlib.pyplot as plt
# Import data
repository = Repository(config)
dataset, labels = repository.get_dataset_and_labels()
dataset = dataset.fillna(-85)
# Data: Accuracy && Error
acc = []
err = []
size = []
#Files: AccuracyDTrea && ErrorDTrea
AccuracyDTrea = open("AccuracyDTrea.txt", "w")
ErrorDTrea = open("ErrorDTrea.txt", "w")
# Iterate accross the PCA dimentionality
for i in range(93, 2094, 100):
print "########################"
print "Iteration number: " + str(i / 100)
size.append(i)
# PCA number of components
pca = PCA(n_components=i)
