# Suppress warnings for a clean output
import warnings; warnings.filterwarnings("ignore")
import numpy as np
np.random.seed(42)

# Data setup
import pandas as pd
url = 'https://raw.githubusercontent.com/uiuc-cse/data-fa14/gh-pages/data/iris.csv'
df = pd.read_csv(url)
print(df.shape)
print(df.head(3))

# What columns are in our dataset?
print(df.columns.tolist())

# Basic info and missing value check
info = df.info()
missing = df.isnull().sum()
print(missing)

# Show summary statistics
print(df.describe())

# Plotting petal length vs petal width
import matplotlib.pyplot as plt
plt.figure(figsize=(6,4))
for species in df['species'].unique():
    plt.scatter(df[df['species'] == species]['petal_length'],
                df[df['species'] == species]['petal_width'], label=species)
plt.xlabel('Petal Length')
plt.ylabel('Petal Width')
plt.title('Petal Size by Flower Species')
plt.legend()
plt.show()

# Histogram of sepal length for all flowers
df['sepal_length'].hist(bins=20, color='skyblue', edgecolor='black')
plt.xlabel('Sepal Length')
plt.ylabel('Frequency')
plt.title('Distribution of Sepal Lengths in Dataset')
plt.show()

# Pairplot for relationships among all features
import seaborn as sns
sns.pairplot(df, hue='species', diag_kind='hist')
plt.show()

# Convert species names to numbers
from sklearn.preprocessing import LabelEncoder
le = LabelEncoder()
df['species_num'] = le.fit_transform(df['species'])
print(df[['species', 'species_num']].head(6))

# Split data into train and test sets (80-20 split)
from sklearn.model_selection import train_test_split
features = df.drop(['species', 'species_num'], axis=1).columns
X = df[features]
y = df['species_num']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
print('Train:', X_train.shape, 'Test:', X_test.shape)

# Fit a KNN classifier
from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier(n_neighbors=5)
knn.fit(X_train, y_train)
score = knn.score(X_test, y_test)
print(f"KNN accuracy: {score:.2%}")

# What features matter most? Try with fewer features
X_train_small = X_train[['petal_length', 'petal_width']]
X_test_small = X_test[['petal_length', 'petal_width']]
knn2 = KNeighborsClassifier(n_neighbors=5)
knn2.fit(X_train_small, y_train)
score2 = knn2.score(X_test_small, y_test)
print(f"KNN with petals only: {score2:.2%}")

# Predict the class of a new flower (interactive)
s_length = float(input("Enter sepal length in cm: "))
s_width = float(input("Enter sepal width in cm: "))
p_length = float(input("Enter petal length in cm: "))
p_width = float(input("Enter petal width in cm: "))
sample = [[s_length, s_width, p_length, p_width]]
pred = knn.predict(sample)[0]
print(f"Predicted species: {le.inverse_transform([pred])[0]}")

# Confusion matrix for detailed performance
from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay
import numpy as np
y_pred = knn.predict(X_test)
cm = confusion_matrix(y_test, y_pred)
disp = ConfusionMatrixDisplay(confusion_matrix=cm, display_labels=le.classes_)
disp.plot(cmap='Blues')
plt.title('KNN Classification Confusion Matrix')
plt.show()

# Can other algorithms do better? Try a decision tree
from sklearn.tree import DecisionTreeClassifier
tree = DecisionTreeClassifier(random_state=42)
tree.fit(X_train, y_train)
tree_score = tree.score(X_test, y_test)
print(f"Decision Tree accuracy: {tree_score:.2%}")

# Visualize the actual tree structure (simplified)
from sklearn import tree as sktree
plt.figure(figsize=(10,6))
sktree.plot_tree(tree, feature_names=features, class_names=le.classes_, filled=True, max_depth=2)
plt.title('Decision Tree Structure (First 2 Levels)')
plt.show()