def change_cwd_data(self):
cwd = os.getcwd()
os.chdir(
DATA_PATH.format(dim=self.__dim,
reference=self.__reference_distribution))
return cwd
from Constants.Constants import SIZE_SET, DIMENSION_SET, SEED
from Constants.Expanded_Constants import REFERENCE_LIST
from Constants.Storage_Constants import DATA_PATH
n_components = []
explained_variance_ratio = []
for dim in DIMENSION_SET:
print('Dimension:', dim)
size_n_components = []
size_explained_var = []
for size in SIZE_SET:
dist_n_components = []
dist_explained_var = []
print('\t' + str(size))
for reference in REFERENCE_LIST:
os.chdir(DATA_PATH.format(dim=dim, reference=reference))
size_files = glob.glob('*Data Set ' + str(size) + '.parquet_' +
reference + '_gz')
feature_columns = ['U' + str(i) for i in range(size + 1)
] + ['V' + str(i) for i in range(size + 1)]
size_df = pd.concat(map(lambda x: pd.read_parquet(x), size_files),
ignore_index=True)[feature_columns]
for percent in [0.85, 0.90, 0.95, 0.99, 0.999]:
print('\t\t' + reference + '\t' + str(percent), end=' \t')
pca = PCA(n_components=percent, random_state=SEED)
pca.fit(size_df)
np_values = pca.transform(size_df)
print(str(pca.n_components_), flush=True)
n_components.append(pca.n_components)
dist_n_components.append(pca.n_components)
print('\t\t\tExplained Variance:',
import glob
import os
import numpy as np
import pandas as pd
from matplotlib import pyplot as plt
from Constants.Constants import SIZE_SET
from Constants.Expanded_Constants import REFERENCE_LIST
from Constants.Storage_Constants import DATA_PATH
for size in SIZE_SET:
print(1, size, 'Complete')
for reference in REFERENCE_LIST:
print('\t', reference, sep='')
os.chdir(DATA_PATH.format(dim=1, reference=reference))
cluster = glob.glob('*Set ' + str(size) + '.parquet_' + reference +
'_gz')
plt.close('all')
for file in cluster:
dist = file.split(' ')[:1]
dist_name = ' '.join(dist)
file_df = pd.read_parquet(file).astype(dtype=np.float32)
print('\t\t', file, sep='')
u = file_df['U'].to_numpy()
v = file_df['V'].to_numpy()
plt.plot(u, v, '.', label=dist_name)
plt.legend()
plt.xlabel('U')
plt.ylabel('V')
plt.title('Dim 1 Size ' + str(size) + ' ' + reference)