kyoto-cheng

# weights for each style layer 
# weighting earlier layers more will result in *larger* style artifacts
# notice we are excluding `conv4_2` our content representation
style_weights = {'conv1_1': 1.,
                 'conv2_1': 0.75,
                 'conv3_1': 0.2,
                 'conv4_1': 0.2,
                 'conv5_1': 0.2}

content_weight = 1  # alpha

kyoto-cheng

def get_features(image, model, layers=None):
    """ Run an image forward through a model and get the features for 
        a set of layers. Default layers are for VGGNet matching Gatys et al (2016)
    """
    
    ## Complete mapping layer names of PyTorch's VGGNet to names from the paper
    ## Need the layers for the content and style representations of an image
    if layers is None:
        layers = {'0': 'conv1_1',
                  '5': 'conv2_1', 

kyoto-cheng

# import resources
%matplotlib inline
from PIL import Image
import matplotlib.pyplot as plt
import numpy as np
import torch
import torch.optim as optim
from torchvision import transforms, models

# get the "features" portion of VGG19 (we will not need the "classifier" portion)

kyoto-cheng

sns.violinplot(x='Credit_Limit', y='Income_Category', data=data, hue="Attrition_Flag", split=True, palette="Set3", scale="width", height=4, aspect=.7).set_title('Original Data')
sns.violinplot(x='Credit_Limit', y='Income_Category', data=new_data_model_CTGAN, hue="Attrition_Flag", split=True, palette="Set3", scale="width", height=4, aspect=.7).set_title('model_CTGAN')
sns.violinplot(x='Credit_Limit', y='Income_Category', data=new_data_model_CopulaGAN, hue="Attrition_Flag", split=True, palette="Set3", scale="width", height=4, aspect=.7).set_title('model_CopulaGAN')

kyoto-cheng

# pip install sdv
# importing the necesary libraries
import numpy as np 
import pandas as pd 
import warnings
warnings.filterwarnings('ignore')

# import all 4 sdv models under the single table scenario
from sdv.tabular import GaussianCopula
from sdv.tabular import CTGAN

kyoto-cheng

# Fit and transform the Vectorizer based on the feature selection results X_names 
vectorizer = feature_extraction.text.CountVectorizer(vocabulary=X_names)
vectorizer.fit(corpus)
X_train = vectorizer.transform(corpus)

# Testing ML models are Naive Bayes, Random Forest and Decision Trees 
NB_Classifier = naive_bayes.MultinomialNB()
RForest_Classifier = RandomForestClassifier()
DTree_Classifier = DecisionTreeClassifier()

kyoto-cheng

import re
import pandas as pd
import numpy as np
import seaborn as sns
from sklearn.feature_extraction.text import CountVectorizer
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn import feature_extraction, model_selection, naive_bayes, pipeline, manifold, preprocessing, feature_selection
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import confusion_matrix

kyoto-cheng

from wordcloud import WordCloud
import matplotlib.pyplot as plt 
from sklearn.feature_extraction.text import CountVectorizer    
    
# Analyzing top frquent bi-gram words in the interview questions of type Methodology with CountVectorizer
def counter(Q_A, category, data, n_gram_min, n_gram_max): 
    data = data[data[category]==1]
    word_vectorizer = CountVectorizer(ngram_range=(n_gram_min,n_gram_max), analyzer='word')
    sparse_matrix = word_vectorizer.fit_transform(data[Q_A])
    frequencies = sum(sparse_matrix).toarray()[0]

kyoto-cheng

import re
import pandas as pd
import nltk
from nltk import word_tokenize

nltk.download('wordnet')
nltk.download('punkt')

lemma = nltk.wordnet.WordNetLemmatizer()

kyoto-cheng

import pandas as pd
import numpy as np
import re
from nltk.corpus import stopwords
import matplotlib.pyplot as plt
from sklearn.base import BaseEstimator, TransformerMixin
from sklearn.model_selection import train_test_split
from sklearn.pipeline import Pipeline
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.preprocessing import StandardScaler