Heatmaps: 2D Matrix Encodings & Correlation Analysis

import pandas as pd
import numpy as np

df = pd.read_csv('ecommerce_sales.csv')

# 1. Pearson Correlation (Default: Linear associations)
# Highly sensitive to extreme outliers; assumes approximate normality
pearson_corr = df.corr(method='pearson', numeric_only=True)

# 2. Spearman Rank Correlation (Monotonic rank-order associations)
# Uses ranks instead of raw values; robust to non-linear curves & extreme outliers
spearman_corr = df.corr(method='spearman', numeric_only=True)

print(pearson_corr.round(2))

import seaborn as sns
import matplotlib.pyplot as plt
import numpy as np

corr = df.corr(numeric_only=True)

fig, ax = plt.subplots(figsize=(10, 8), dpi=100)

sns.heatmap(
    corr,
    cmap='coolwarm',       # Diverging palette (blue = negative, red = positive)
    vmin=-1.0,             # CRITICAL: Anchors scale floor (prevents false saturation)
    vmax=1.0,              # CRITICAL: Anchors scale ceiling
    center=0.0,            # Neutral white/gray exactly at zero correlation
    annot=True,            # Prints numerical coefficients inside cells
    fmt='.2f',             # Clean 2-decimal formatting
    square=True,           # Forces 1:1 square cells (industry visual standard)
    linewidths=0.75,       # Visual gap separating adjacent squares
    linecolor='#0f172a',   # Dark separator color matching theme
    cbar_kws={'label': 'Pearson Correlation (r)', 'shrink': 0.8},
    ax=ax
)

ax.set_title("Feature Correlation Matrix", fontsize=14, fontweight='bold', pad=14)
plt.tight_layout()
plt.show()

import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt

corr = df.corr()

# 1. Generate boolean array of identical shape
# 2. np.triu() sets the upper triangle + diagonal to True
mask = np.triu(np.ones_like(corr, dtype=bool))

fig, ax = plt.subplots(figsize=(8, 6))

# Seaborn hides any cell where mask is True!
sns.heatmap(
    corr,
    mask=mask,
    cmap='coolwarm',
    vmin=-1.0,
    vmax=1.0,
    center=0,
    annot=True,
    square=True,
    ax=ax
)
plt.show()

import seaborn as sns
import matplotlib.pyplot as plt
import pandas as pd

# 1. Pivot raw logs into a 7 x 24 matrix
pivot = df.pivot_table(
    index='day_of_week',
    columns='hour_of_day',
    values='request_id',
    aggfunc='count',
    fill_value=0
)

fig, ax = plt.subplots(figsize=(12, 5), dpi=100)

# 2. Sequential palette for counts (YlOrRd: Light Yellow to Deep Red)
sns.heatmap(
    pivot,
    cmap='YlOrRd',
    linewidths=0.5,
    linecolor='#0f172a',
    cbar_kws={'label': 'Visitor Count'},
    ax=ax
)

ax.set_title("24x7 Web Traffic Load Heatmap", fontweight='bold')
plt.show()

import seaborn as sns
import matplotlib.pyplot as plt

# Clustermap returns a specialized ClusterGrid object
g = sns.clustermap(
    df.corr(),
    method='ward',              # Ward variance minimization linkage
    metric='euclidean',         # Geometric distance
    cmap='vlag',
    vmin=-1.0,
    vmax=1.0,
    center=0,
    annot=True,
    figsize=(9, 9),
    dendrogram_ratio=(0.18, 0.18) # Fraction of canvas reserved for tree branches
)

g.fig.suptitle("Feature Hierarchical Clustering", y=1.02, fontweight='bold')
plt.show()

import matplotlib.pyplot as plt
import numpy as np

data_matrix = np.random.rand(6, 6)

fig, ax = plt.subplots(figsize=(7, 6))

# ax.imshow renders raw 2D image pixels
im = ax.imshow(data_matrix, cmap='coolwarm', vmin=0, vmax=1, aspect='equal')

# Manually attach colorbar
cbar = fig.colorbar(im, ax=ax, shrink=0.8)
cbar.set_label('Normalized Intensity')

ax.set_xticks(range(6))
ax.set_yticks(range(6))
ax.set_title("Native Matplotlib ax.imshow() Matrix", fontweight='bold')
plt.show()