The Iris dataset is one of the most well-known datasets in machine learning. It contains 150 samples from three species of Iris flowers (Iris setosa, Iris versicolor, and Iris virginica), with four features measured for each sample: sepal length, sepal width, petal length, and petal width.
In this project, we perform an exploratory data analysis (EDA) to understand the distribution of features, relationships between variables, and build a simple classification model.
df.head(10)| sepal length (cm) | sepal width (cm) | petal length (cm) | petal width (cm) | species | |
|---|---|---|---|---|---|
| 0 | 5.1 | 3.5 | 1.4 | 0.2 | setosa |
| 1 | 4.9 | 3.0 | 1.4 | 0.2 | setosa |
| 2 | 4.7 | 3.2 | 1.3 | 0.2 | setosa |
| 3 | 4.6 | 3.1 | 1.5 | 0.2 | setosa |
| 4 | 5.0 | 3.6 | 1.4 | 0.2 | setosa |
| 5 | 5.4 | 3.9 | 1.7 | 0.4 | setosa |
| 6 | 4.6 | 3.4 | 1.4 | 0.3 | setosa |
| 7 | 5.0 | 3.4 | 1.5 | 0.2 | setosa |
| 8 | 4.4 | 2.9 | 1.4 | 0.2 | setosa |
| 9 | 4.9 | 3.1 | 1.5 | 0.1 | setosa |
The dataset has 150 samples across 3 species, each with 4 numerical features.
df.groupby('species').describe().T.round(2)/tmp/ipykernel_348982/2432499185.py:1: FutureWarning:
The default of observed=False is deprecated and will be changed to True in a future version of pandas. Pass observed=False to retain current behavior or observed=True to adopt the future default and silence this warning.
| species | setosa | versicolor | virginica | |
|---|---|---|---|---|
| sepal length (cm) | count | 50.00 | 50.00 | 50.00 |
| mean | 5.01 | 5.94 | 6.59 | |
| std | 0.35 | 0.52 | 0.64 | |
| min | 4.30 | 4.90 | 4.90 | |
| 25% | 4.80 | 5.60 | 6.22 | |
| 50% | 5.00 | 5.90 | 6.50 | |
| 75% | 5.20 | 6.30 | 6.90 | |
| max | 5.80 | 7.00 | 7.90 | |
| sepal width (cm) | count | 50.00 | 50.00 | 50.00 |
| mean | 3.43 | 2.77 | 2.97 | |
| std | 0.38 | 0.31 | 0.32 | |
| min | 2.30 | 2.00 | 2.20 | |
| 25% | 3.20 | 2.52 | 2.80 | |
| 50% | 3.40 | 2.80 | 3.00 | |
| 75% | 3.68 | 3.00 | 3.18 | |
| max | 4.40 | 3.40 | 3.80 | |
| petal length (cm) | count | 50.00 | 50.00 | 50.00 |
| mean | 1.46 | 4.26 | 5.55 | |
| std | 0.17 | 0.47 | 0.55 | |
| min | 1.00 | 3.00 | 4.50 | |
| 25% | 1.40 | 4.00 | 5.10 | |
| 50% | 1.50 | 4.35 | 5.55 | |
| 75% | 1.58 | 4.60 | 5.88 | |
| max | 1.90 | 5.10 | 6.90 | |
| petal width (cm) | count | 50.00 | 50.00 | 50.00 |
| mean | 0.25 | 1.33 | 2.03 | |
| std | 0.11 | 0.20 | 0.27 | |
| min | 0.10 | 1.00 | 1.40 | |
| 25% | 0.20 | 1.20 | 1.80 | |
| 50% | 0.20 | 1.30 | 2.00 | |
| 75% | 0.30 | 1.50 | 2.30 | |
| max | 0.60 | 1.80 | 2.50 |
Violin plots reveal how each feature is distributed across the three Iris species. Notice how petal measurements provide much clearer separation between species compared to sepal measurements.
fig, axes = plt.subplots(2, 2, figsize=(12, 9))
features = iris.feature_names
colors = ['#63b3ed', '#68d391', '#fc8181']
for idx, (ax, feature) in enumerate(zip(axes.flat, features)):
sns.violinplot(
data=df, x='species', y=feature, ax=ax,
palette=colors, inner='box', linewidth=1.2
)
ax.set_title(feature.replace(' (cm)', '').title(), fontweight='bold')
ax.set_xlabel('')
ax.set_ylabel('cm')
plt.tight_layout()
plt.savefig('iris-violin.png', dpi=150, bbox_inches='tight', facecolor='#1b2838')/tmp/ipykernel_348982/4012254396.py:6: FutureWarning:
Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.
/tmp/ipykernel_348982/4012254396.py:6: FutureWarning:
Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.
/tmp/ipykernel_348982/4012254396.py:6: FutureWarning:
Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.
/tmp/ipykernel_348982/4012254396.py:6: FutureWarning:
Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.
A pairplot shows all feature combinations, revealing clusters and the separability of each species.
palette = {'setosa': '#63b3ed', 'versicolor': '#68d391', 'virginica': '#fc8181'}
g = sns.pairplot(
df, hue='species', palette=palette,
diag_kind='kde', plot_kws={'alpha': 0.7, 's': 40, 'edgecolor': 'white', 'linewidth': 0.3},
diag_kws={'linewidth': 2}
)
g.figure.set_facecolor('#1b2838')
for ax in g.axes.flat:
ax.set_facecolor('#0d1b2a')
ax.tick_params(colors='#94a3b8')
ax.xaxis.label.set_color('#cbd5e1')
ax.yaxis.label.set_color('#cbd5e1')
g.legend.get_frame().set_facecolor('#1b2838')
g.legend.get_frame().set_edgecolor('#4a5568')
for text in g.legend.get_texts():
text.set_color('#cbd5e1')
plt.savefig('iris-pairplot.png', dpi=150, bbox_inches='tight', facecolor='#1b2838')
Key observations:
fig, ax = plt.subplots(figsize=(8, 6))
corr = df.drop('species', axis=1).corr()
mask = np.triu(np.ones_like(corr, dtype=bool))
cmap = sns.diverging_palette(220, 20, as_cmap=True)
sns.heatmap(
corr, mask=mask, annot=True, fmt='.2f', cmap=cmap,
center=0, square=True, linewidths=2, linecolor='#2d3748',
cbar_kws={'shrink': 0.8, 'label': 'Correlation'},
ax=ax, vmin=-1, vmax=1,
annot_kws={'size': 13, 'weight': 'bold'}
)
ax.set_title('Feature Correlation Matrix', fontweight='bold', pad=15)
labels = [name.replace(' (cm)', '').title() for name in iris.feature_names]
ax.set_xticklabels(labels, rotation=45, ha='right')
ax.set_yticklabels(labels, rotation=0)
plt.tight_layout()
plt.savefig('iris-heatmap.png', dpi=150, bbox_inches='tight', facecolor='#1b2838')
Petal length and petal width are highly correlated (r = 0.96), suggesting they capture similar information about the flower’s structure.
We train a Random Forest classifier to predict species from the four features.
# Split data
X_train, X_test, y_train, y_test = train_test_split(
iris.data, iris.target, test_size=0.3, random_state=42, stratify=iris.target
)
# Train model
clf = RandomForestClassifier(n_estimators=100, random_state=42)
clf.fit(X_train, y_train)
# Predict
y_pred = clf.predict(X_test)
accuracy = accuracy_score(y_test, y_pred)
# Confusion matrix
cm = confusion_matrix(y_test, y_pred)
fig, ax = plt.subplots(figsize=(7, 6))
sns.heatmap(
cm, annot=True, fmt='d', cmap='Blues',
xticklabels=iris.target_names,
yticklabels=iris.target_names,
linewidths=2, linecolor='#2d3748',
cbar=False, ax=ax,
annot_kws={'size': 18, 'weight': 'bold'}
)
ax.set_xlabel('Predicted', fontweight='bold')
ax.set_ylabel('Actual', fontweight='bold')
ax.set_title(f'Random Forest — Accuracy: {accuracy:.1%}', fontweight='bold', pad=15)
plt.tight_layout()
plt.savefig('iris-confusion.png', dpi=150, bbox_inches='tight', facecolor='#1b2838')
importances = clf.feature_importances_
feature_labels = [name.replace(' (cm)', '').title() for name in iris.feature_names]
sorted_idx = np.argsort(importances)
fig, ax = plt.subplots(figsize=(8, 5))
colors = ['#63b3ed' if i >= 2 else '#4a5568' for i in sorted_idx]
ax.barh(range(len(sorted_idx)), importances[sorted_idx], color=colors, edgecolor='#2d3748', height=0.6)
ax.set_yticks(range(len(sorted_idx)))
ax.set_yticklabels([feature_labels[i] for i in sorted_idx])
ax.set_xlabel('Importance', fontweight='bold')
ax.set_title('Feature Importance (Random Forest)', fontweight='bold', pad=15)
for i, v in enumerate(importances[sorted_idx]):
ax.text(v + 0.01, i, f'{v:.3f}', va='center', color='#cbd5e1', fontweight='bold')
plt.tight_layout()
plt.savefig('iris-importance.png', dpi=150, bbox_inches='tight', facecolor='#1b2838')
| Aspect | Finding |
|---|---|
| Best separators | Petal length and petal width |
| Easiest species | Setosa — linearly separable |
| Hardest pair | Versicolor vs Virginica |
| Model accuracy | Random Forest achieves near-perfect accuracy on this dataset |
| Top features | Petal width and petal length dominate feature importance |
This analysis demonstrates a standard EDA workflow: understand the data structure, visualize distributions, explore correlations, and build a baseline model.