Project 2

1. Display the data of white wine and red wine:

• Assuming the dataset in a pandas DataFrame (df), filter the white and red wines as follows:

  white_wine = df[df['type'] == 'white']
  red_wine = df[df['type'] == 'red']
  print(white_wine)
  print(red_wine)
  

2. Get the size of both white wine and red wine:

• get the size of each dataset by:

  white_size = white_wine.shape[0]
  red_size = red_wine.shape[0]
  print(f"White wine size: {white_size}")
  print(f"Red wine size: {red_size}")
  

3. Get the number of nulls in both types of wines:

• To find the number of null values:

  white_nulls = white_wine.isnull().sum()
  red_nulls = red_wine.isnull().sum()
  print(f"White wine nulls:\n{white_nulls}")
  print(f"Red wine nulls:\n{red_nulls}")
  

4. Check for duplicates and remove them in both types of wine data:

• To check for and remove duplicates:

  white_wine = white_wine.drop_duplicates()
  red_wine = red_wine.drop_duplicates()
  

5. Get the number of unique values in all features and print them:

• To check unique values:

  white_uniques = white_wine.nunique()
  red_uniques = red_wine.nunique()
  print(f"Unique values in white wine:\n{white_uniques}")
  print(f"Unique values in red wine:\n{red_uniques}")
  

6. Print the mean density of both the sets:

• To calculate and print the mean density:

  mean_density_white = white_wine['density'].mean()
  mean_density_red = red_wine['density'].mean()
  print(f"Mean density of white wine: {mean_density_white}")
  print(f"Mean density of red wine: {mean_density_red}")
  

7. Combine both datasets with the color of the wine in a column:

• To combine the datasets and add a 'color' column:

  white_wine['color'] = 'white'
  red_wine['color'] = 'red'
  combined_wine = pd.concat([white_wine, red_wine])
  

8. Remove the null values if any:

• To remove rows with null values:

  combined_wine = combined_wine.dropna()
  

9. Remove the outliers if any:

• You can use the Z-score method or IQR method to detect and remove outliers. For simplicity, here's an IQR-based method:

  Q1 = combined_wine['density'].quantile(0.25)
  Q3 = combined_wine['density'].quantile(0.75)
  IQR = Q3 - Q1
  combined_wine = combined_wine[(combined_wine['density'] >= (Q1 - 1.5 * IQR)) & (combined_wine['density'] <= (Q3 + 1.5 * IQR))]
  

10. Do wines with higher alcohol content receive better ratings?

• Use a correlation between alcohol content and quality:

  correlation = combined_wine['alcohol'].corr(combined_wine['quality'])
  print(f"Correlation between alcohol content and quality: {correlation}")
  

11. Find the correlation between citric acid content and pH:

• To find the correlation:

  correlation_citric_ph = combined_wine['citric acid'].corr(combined_wine['pH'])
  print(f"Correlation between citric acid and pH: {correlation_citric_ph}")
  

12. Find regression of pH on citric acid:

• To perform regression:

  from sklearn.linear_model import LinearRegression
  X = combined_wine[['citric acid']]
  y = combined_wine['pH']
  
  model = LinearRegression()
  model.fit(X, y)
  print(f"Regression Coefficients: {model.coef_}, Intercept: {model.intercept_}")
  

13. Find how many variables are independent:

• To check multicollinearity and find independent variables, calculate the Variance Inflation Factor (VIF) for each feature:

  from statsmodels.stats.outliers_influence import variance_inflation_factor
  X = combined_wine.drop(columns='quality')
  vif = pd.DataFrame()
  vif['Variable'] = X.columns
  vif['VIF'] = [variance_inflation_factor(X.values, i) for i in range(len(X.columns))]
  print(vif)
  

14. Create a new column as the acidity level. Divide the existing data into 5 groups based on the acidity level. The levels are to be named ['High', 'Moderately_High', 'Medium', 'Low']. Acidity may be obtained from the pH.

• To create the acidity level:

  bins = [0, 3.2, 3.5, 3.8, 4.0, 4.5]
  labels = ['High', 'Moderately_High', 'Medium', 'Low', 'Very_Low']
  combined_wine['acidity_level'] = pd.cut(combined_wine['pH'], bins=bins, labels=labels)
  

15. Find the mean quality of each acidity level with groupby. Use groupby:

• To calculate mean quality for each acidity level:

  mean_quality_acidity = combined_wine.groupby('acidity_level')['quality'].mean()
  print(f"Mean quality for each acidity level:\n{mean_quality_acidity}")
  

16. Is a certain type of wine (red or white) associated with higher quality?

• To check the average quality of each wine type:

  mean_quality_wine_type = combined_wine.groupby('color')['quality'].mean()
  print(f"Mean quality for each wine type:\n{mean_quality_wine_type}")
  

17. Select samples with alcohol content less than the median:

• To filter the samples:

  median_alcohol = combined_wine['alcohol'].median()
  low_alcohol = combined_wine[combined_wine['alcohol'] < median_alcohol]
  print(f"Samples with alcohol content less than the median:\n{low_alcohol}")
  

18. Get the quality rating for the low alcohol and high alcohol groups:

• To separate by alcohol content:

  high_alcohol = combined_wine[combined_wine['alcohol'] >= median_alcohol]
  print(f"Low alcohol quality ratings:\n{low_alcohol['quality']}")
  print(f"High alcohol quality ratings:\n{high_alcohol['quality']}")
  

19. Do sweeter wines get more ratings?

• Check if residual sugar correlates with quality:

  correlation_sugar_quality = combined_wine['residual sugar'].corr(combined_wine['quality'])
  print(f"Correlation between residual sugar and quality: {correlation_sugar_quality}")
  

20. Get the number of counts based on ratings and color of wine:

• To count by wine color and ratings:

  counts_by_rating_color = combined_wine.groupby(['color', 'quality']).size()
  print(f"Counts based on ratings and wine color:\n{counts_by_rating_color}")
  

21. Is the density of red wine and white wine similar?

• To compare density of red and white wines:

  density_comparison = combined_wine.groupby('color')['density'].mean()
  print(f"Average density for each wine type:\n{density_comparison}")
  

22. Is the confidence interval for the differences of the proportion of white wine with rating 3 and the proportion of red wine with rating 3?

• Perform a Z-test for proportions to compare the proportions of white and red wines with a rating of 3:

  from statsmodels.stats.proportion import proportions_ztest
  white_rating_3 = (white_wine['quality'] == 3).sum()
  red_rating_3 = (red_wine['quality'] == 3).sum()
  
  n_white = white_wine.shape[0]
  n_red = red_wine.shape[0]
  
  z_stat, p_value = proportions_ztest([white_rating_3, red_rating_3], [n_white, n_red])
  print(f"Z-statistic: {z_stat}, p-value: {p_value}")