Project 2

Q1

wine_red=pd.read_csv('winequality-red.csv')

wine_red.head()

wine_white=pd.read_csv('winequality-white.csv')

wine_white.head()

Q2.

wine_white.info()

wine_red.info()

Q3

wine_red.isnull().sum()

wine_white.isnull().sum()

Q4

wine_red[wine_red.duplicated()]

wine_red.drop_duplicates()

wine_white[wine_white.duplicated()]

wine_white.drop_duplicates()

Q5

wine_red.nunique()

wine_white.nunique()

Q6

Q7

Q8

df_wine.isnull().sum()

df_wine[df_wine.duplicated()]

df_wine.drop_duplicates()

Q9

df_wine['color'] = df_wine['color'].infer_objects(copy=False)

# Remove outliers for each numerical column

for col in df_wine.select_dtypes(include=[np.number]).columns:

Q1 = df_wine[col].quantile(0.25)

Q3 = df_wine[col].quantile(0.75)

IQR = Q3 - Q1

lower_bound = Q1 - 1.5 * IQR

upper_bound = Q3 + 1.5 * IQR

df_wine = df_wine[(df_wine[col] >= lower_bound) & (df_wine[col] <= upper_bound)]

# Display the cleaned DataFrame

print("Cleaned DataFrame without outliers:")

df_wine

Q10

columns_to_check = ['fixed acidity','volatile acidity','citric acid','residual sugar','chlorides','free sulfur dioxide','density','pH','sulphates','alcohol','quality']

# Calculate the Z-scores for each data point in the specified columns

z_scores = np.abs((df_wine[columns_to_check] - df_wine[columns_to_check].mean()) / df_wine[columns_to_check].std())

# Set the threshold for outlier detection (Z-score greater than 3)

threshold = 3

# Remove the rows with outliers

df_wine[(z_scores < threshold)]

correlation=df_wine['alcohol'].corr(df_wine['quality'])

print(correlation)

# Plot a scatter plot to visualize the relationship

sns.scatterplot(x='alcohol', y='quality', data=df_wine)

# Display the plot

plt.show()

Q11

correlation=df_wine['citric acid'].corr(df_wine['pH'])

print(correlation)

Q12

from sklearn.linear_model import LinearRegression

# Select the predictor (citric acid) and target variable (pH)

X = df_wine[['citric acid']]

y = df_wine['pH']

# Create an instance of the LinearRegression model

model = LinearRegression()

# Fit the model to the data

model.fit(X, y)

# Get the regression coefficients

intercept = model.intercept_

slope = model.coef_[0]

print("Regression equation : pH = {:.2f} + {:.2f} * citric acid ".format(intercept,slope))

plt.figure(figsize=(10, 6))

plt.scatter(X, y, color='blue', label='Data points')

plt.plot(X, model.predict(X), color='red', linewidth=2, label='Regression line')

plt.xlabel('Citric Acid')

plt.ylabel('pH')

plt.title('Regression of pH on Citric Acid')

plt.legend()

plt.show()

Q13

from scipy.stats import chi2_contingency

# Create a contingency table

contingency_table = pd.crosstab(df_wine["citric acid"], df_wine["pH"])

# Calculate the chi-square statistic

chi_square_statistic, p_value, df, _ = chi2_contingency(contingency_table)

# Print the results

print("Chi-square statistic:", chi_square_statistic)

print("P-value:", p_value)

# Interpretation

if p_value < 0.05:

print("The variables are statistically dependent.")

else:

print("The variables are statistically independent.")

Q14

bin_edges = [2.72,3.11,3.21,3.32,4.01]

bin_names = ['very_high','high','medium','low']

df_wine['acidity_levels'] = pd.cut(df_wine['pH'], bin_edges, labels=bin_names)

df_wine.head()

Q15

df_wine.groupby('acidity_levels',observed=True)['quality'].mean()

Q16

drdf.groupby('color',observed=True)['quality'].mean()

Q17

#low_alcohol = df_wine[df_wine['alcohol'] < 10.3]

low_alcohol=df_wine.query('alcohol < 10.3')

low_alcohol

#high_alcohol = df_wine[df_wine['alcohol'] >= 10.3]

high_alcohol =df_wine.query('alcohol >= 10.3')

high_alcohol

Q18

Q19

correlation = df_wine['residual sugar'].corr(df_wine['quality'])

print(correlation)

# visualize the relationship between sweetness level and wine ratings using a scatter plot

import matplotlib.pyplot as plt

plt.scatter(df_wine['residual sugar'], df_wine['quality'])

plt.xlabel('Residual Sugar')

plt.ylabel('Quality')

plt.title('Relationship between Sweetness and Wine Quality')

plt.show()

Q20

# Count the number of occurrences based on ratings and color

count_by_ratings_and_color = df_wine.groupby(['quality', 'color']).size()

# Print the counts

print(count_by_ratings_and_color)

Q21

from scipy import stats

# Extract the density for red and white wines

red_wine_density = df_wine.loc[df_wine['color'] == 'red', 'density']

white_wine_density = df_wine.loc[df_wine['color'] == 'white', 'density']

# Perform a t-test

t_statistic, p_value = stats.ttest_ind(red_wine_density, white_wine_density, equal_var=False)

# Compare the p-value to the significance level (e.g., 0.05)

alpha = 0.05

if p_value < alpha:

print("The density of red wine and white wine is significantly different.")

else:

print("The density of red wine and white wine is not significantly different.")

Q22

from math import sqrt

# Replace the placeholders with the actual counts and totals from your dataset

count_white_rating3 = 50 # Replace with the count of white wines with rating 3

total_white_wines = 200 # Replace with the total count of white wines

count_red_rating3 = 30 # Replace with the count of red wines with rating 3

total_red_wines = 150 # Replace with the total count of red wines

# Calculate the proportions for white wine with rating 3 and red wine with rating 3

proportion1 = count_white_rating3 / total_white_wines

proportion2 = count_red_rating3 / total_red_wines

# Calculate the sample sizes

n1 = total_white_wines

n2 = total_red_wines

standard_error = sqrt((proportion1 * (1 - proportion1) / n1) + (proportion2 * (1 - proportion2) / n2))

from scipy.stats import norm

# Set the desired confidence level (e.g., 95%)

confidence_level = 0.95

# Calculate the critical value (z-score) using the norm function from scipy.stats

critical_value = norm.ppf((1 + confidence_level) / 2)

# Calculate the margin of error

margin_of_error = critical_value * standard_error

# Calculate the lower and upper bounds of the confidence interval

lower_bound = (proportion1 - proportion2) - margin_of_error

upper_bound = (proportion1 - proportion2) + margin_of_error