gvyshnya

### we bin the following numeric variables using gaussian mixture models
bin_these = {'cont0': 4, 'cont1': 5, 'cont3': 2, 'cont4': 2, 'cont6': 3, 'cont8': 3, 'cont10': 10}
train, test = FW.FE_discretize_numeric_variables(train,bin_these,test=test, strategy='gaussian')

gvyshnya

intxn_vars = [('cont3', 'cont7'),('cont3', 'cont8'),('cont3', 'cont9'),('cont3', 'cont10'),('cont4', 'cont5'),
             ('cont4', 'cont6'),('cont4', 'cont9'),('cont4', 'cont10')]

def FE_create_interaction_vars(df, intxn_vars):
    """
    This handy function creates interaction variables among pairs of numeric vars you send in.
    Your input must be a dataframe and a list of tuples. Each tuple must contain a pair of variables.
    All variables must be numeric. Double check your input before sending them in.
    """
    df = df.copy(deep=True)

gvyshnya

from featurewiz import featurewiz

# out1 and out2: is always a tuple containing two objects. 
# In our case, it is going to be "test" and "train" dataframes with the important features selected by featurewiz
trainm, testm = featurewiz(
    df_train, 
    target='target', 
    corr_limit=0.7, 
    verbose=2, 
    sep=",",

gvyshnya

from autoviz.AutoViz_Class import AutoViz_Class

AV = AutoViz_Class()
dftc = AV.AutoViz(
    filename='', 
    sep='' , 
    depVar='target', 
    dfte=df_train, 
    header=0, 
    verbose=2, 

gvyshnya

from autoviz.AutoViz_Class import AutoViz_Class

AV = AutoViz_Class()
dft = AV.AutoViz(filename='', 
                 sep='' , 
                 depVar='cp_type', 
                 dfte=training_gcols, 
                 header=0, 
                 verbose=2, 
                 lowess=False, 

gvyshnya

import sweetviz as sv

df1 = sv.compare(df_tr, df_cntr, pairwise_analysis='on')
df1.show_html(cells_tr_cntr_comparison_path)

gvyshnya

import sweetviz as sv

df1 = sv.analyze(final_training_df, target_label, pairwise_analysis='on')
df1.show_html(sv_report_path)

gvyshnya

def visualize_features_vs_target_label(df_data, label, feature_list, n_cols=3):
    
    if len(feature_list) % n_cols == 0:
    
        number_of_rows = int(len(feature_list)/n_cols)
    else:
        number_of_rows = int(len(feature_list)/n_cols) +1
    
    fig = make_subplots(rows=number_of_rows, cols=n_cols)
    

gvyshnya

import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import datetime as dt
from typing import Tuple

import matplotlib.pyplot as plt
import seaborn as sns
import plotly.graph_objects as go
from plotly.subplots import make_subplots
import plotly.express as px

gvyshnya

import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import datetime as dt
from typing import Tuple

import matplotlib.pyplot as plt
import seaborn as sns
import plotly.graph_objects as go
from plotly.subplots import make_subplots
import plotly.express as px