grahamharrison68

def normal_distribution_ci(confidence, x_bar, sigma, n):
    z_score = stats.norm.interval(confidence)[1]
    sigma_over_root_n = sigma / np.sqrt(n)
    ci = [x_bar - z_score * sigma_over_root_n, x_bar + z_score * sigma_over_root_n]
    return ci

def binomial_distribution_ci(confidence, p_hat, n):
    z_score = stats.norm.interval(confidence)[1]
    rhs = z_score * np.sqrt(p_hat*(1-p_hat))/n
    ci = [p_hat - rhs, p_hat + rhs]

grahamharrison68

from ipywidgets import interact, interactive, fixed, interact_manual
import ipywidgets as widgets
import scipy.stats as stats
import numpy as np
import pandas as pd

pd.set_option("max_colwidth", 150)

grahamharrison68

df_wine_balanced['label'].value_counts(normalize=True)

grahamharrison68

red_wines = df_merged_wine[df_merged_wine['label'] == 1]
all_white_wines = df_merged_wine[df_merged_wine['label'] == 0]

white_wines = all_white_wines.sample(n=red_wines.shape[0], random_state=24)

df_wine_balanced = pd.concat([red_wines, white_wines])

df_wine_balanced

grahamharrison68

df_merged_wine['label'].value_counts(normalize=True)

grahamharrison68

import pandas as pd
import numpy as np
import graphviz
import pydotplus
import matplotlib.image as mpimg
import io
import random

from matplotlib import pyplot as plt
from sklearn.tree import DecisionTreeClassifier, DecisionTreeRegressor

grahamharrison68

import pandas as pd
import numpy as np
import graphviz
import pydotplus
import matplotlib.image as mpimg
import io
import random

from matplotlib import pyplot as plt
from sklearn.tree import DecisionTreeClassifier, DecisionTreeRegressor

grahamharrison68

# %% Remove the outliers and re-display the box-and-whisker and the histogram
df_normal = remove_all_outliers(df_normal, 'Col0')
box_and_whisker(df_normal, 'Col0')
df_normal['Col0'].hist()

grahamharrison68

box_and_whisker(df_normal, 'Col0')

grahamharrison68

# %% Add some outliers and re-plot 
s[600] = 6
s[700] = 6.5
s[800] = 6.57
s[900] = 6.8

df_normal = pd.DataFrame({'Col0': s})
df_normal['Col0'].hist()