Kansrekenen

import os

import kagglehub
import numpy as np
import pandas as pd
from scipy import stats

# Download data from kaggle
path = kagglehub.dataset_download("yashdevladdha/uber-ride-analytics-dashboard")

# Load data into Pandas DataFrame
csv_file = os.path.join(path, "ncr_ride_bookings.csv")
df = pd.read_csv(csv_file)
print("✅ Data loaded successfully!")

pmf_dict = df["Payment Method"].value_counts(normalize=True).to_dict()

print(pmf_dict)


def pmf(payment_method):
    return pmf_dict.get(payment_method)

pmf("Cash") + pmf("Credit Card") + pmf("Debit Card")

a, b = -3.1, 5.5
interval_start, interval_end = -0.1, 0.6
p_interval_manual = (interval_end - interval_start) / (b - a)
print(p_interval_manual)

normal = stats.norm(loc=0.6, scale=2.2)

# 1. P(x ≤ 0)
p1 = normal.cdf(0)
print(f"1. P(x ≤ 0) = {p1:.4f}")

# 2. P(-1 ≤ x ≤ 1)
p2 = normal.cdf(1) - normal.cdf(-1)
print(f"2. P(-1 ≤ x ≤ 1) = {p2:.4f}")

# 3. P(x > 2)
p3 = 1 - normal.cdf(2)
print(f"3. P(x > 2) = {p3:.4f}")

pmf_dict = df[["Payment Method", "Vehicle Type"]].value_counts(normalize=True).to_dict()

print(pmf_dict)


def pmf(payment_method, vehicle_type):
    return pmf_dict.get((payment_method, vehicle_type))

p1 = pmf("Debit Card", "Bike")
print(f"P(Debit Card, Bike) = {p1:.4f}")

p2 = pmf("Cash", "Uber XL") + pmf("Cash", "Premier Sedan")
print(f"P(Cash, Uber XL or Premier Sedan) = {p2:.4f}")

p3 = sum([pmf(method, "eBike") for method in df["Payment Method"].dropna().unique()])
print(f"P(eBike) = {p3:.4f}")

marginal_ps = {
    method: sum(pmf(method, vehicle_type) for vehicle_type in df["Vehicle Type"].dropna().unique())
    for method in df["Payment Method"].dropna().unique()
}

conditional_ps = {
    method: sum(pmf(method, vehicle_type) for vehicle_type in ["eBike", "Uber XL"])
    for method in df["Payment Method"].dropna().unique()
}

for method, p_marginal in marginal_ps.items():
    print(f"{method}:")
    print(f"\tmarginal: {p_marginal:.4f}")
    p_conditional = conditional_ps.get(method)
    print(f"\tconditional: {p_conditional:.4f}")

pmf_dict = df["Vehicle Type"].value_counts(normalize=True).to_dict()

n = 30
n_xl = int(np.round(pmf_dict.get("Uber XL") * n))
n_eb = int(np.round(pmf_dict.get("eBike") * n))

print(f"# Uber XL: {n_xl}")
print(f"# eBike: {n_eb}")

mu = 500
sigma = 25

for x in [450, 530, 505, 600]:
    p_right = 1 - stats.norm.cdf(x, mu, sigma)
    p_left = stats.norm.cdf(x, mu, sigma)

    if x < mu:
        print(f"x = {x:3d}: P(X≤x) = {p_left:.4f}")
    else:
        print(f"x = {x:3d}: P(X≥x) = {p_right:.4f}")

# Data
batt = np.array(
    [
        23.2,
        25.1,
        22.8,
        24.5,
        23.7,
        25.3,
        22.5,
        26.1,
        24.0,
        23.5,
        24.8,
        23.1,
        25.0,
        24.2,
        23.9,
        22.7,
        25.5,
        24.1,
        23.6,
        24.7,
        23.4,
        24.9,
    ]
)
mu_0 = 24  # Hypothesized mean

# Calculate sample statistics
sample_mean = batt.mean()
sample_std = batt.std(ddof=1)  # N-1 for sample standard deviation
n = len(batt)
se = sample_std / np.sqrt(n)  # Standard error

# Calculate t-statistic manually
t_statistic_manual = (sample_mean - mu_0) / se

# Degrees of freedom
df = n - 1

# Calculate p-value (two-tailed test)
p_value_manual = 2 * (1 - stats.t.cdf(abs(t_statistic_manual), df))

print("Manual")
print(f"t-statistic: {t_statistic_manual:.4f}")
print(f"p-value (two-tailed): {p_value_manual:.4f}")

# Data
control = np.array([72, 68, 75, 71, 69, 74, 70, 73, 68, 71, 72, 70, 74, 69, 71])
treatment = np.array([78, 82, 76, 80, 79, 81, 77, 83, 79, 80, 78, 81, 82, 79, 80])

# Manual calculation
n1 = len(control)
n2 = len(treatment)
mean1 = control.mean()
mean2 = treatment.mean()
std1 = control.std(ddof=1)
std2 = treatment.std(ddof=1)

# Pooled standard deviation (assuming equal variances)
pooled_std = np.sqrt(((n1 - 1) * std1**2 + (n2 - 1) * std2**2) / (n1 + n2 - 2))

# Standard error of difference
se_diff = pooled_std * np.sqrt(1 / n1 + 1 / n2)

# t-statistic
t_stat_manual = (mean1 - mean2) / se_diff

# Degrees of freedom
df_manual = n1 + n2 - 2

# p-value for one-sided test (control > treatment)
p_value_manual = 1 - stats.t.cdf(t_stat_manual, df_manual)

print("Manual")
print(f"t-statistic: {t_stat_manual:.4f}")
print(f"p-value: {p_value_manual:.4f}")