Distribution Charts in Seaborn: Complete Guide

Data analysis ka pehla step hai distribution samajhna — data kaise spread hai, skewed hai ya normal hai, outliers kahan hain. Seaborn ke 6 powerful distribution charts se yeh sab ek glance mein dikhta hai.

📑 Is Guide Mein 6 Distribution Charts Cover Honge:

• sns.histplot() — Enhanced Histogram with KDE overlay
• sns.kdeplot() — Smooth Kernel Density Estimation curve
• sns.rugplot() — Individual data point markers on axis
• sns.ecdfplot() — Empirical Cumulative Distribution
• sns.displot() — Figure-level unified distribution plot
• Combined Distribution Analysis — Hist + KDE + Rug together

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

# Sample Dataset — Employee Data
np.random.seed(42)
df = pd.DataFrame({
    "Salary": np.concatenate([
        np.random.normal(55000, 10000, 400),   # HR/Ops cluster
        np.random.normal(82000, 15000, 350),   # IT/Finance cluster
        np.random.normal(120000, 20000, 150)   # Senior Management
    ]),
    "Department": np.random.choice(["IT","HR","Finance","Marketing"], 900),
    "Experience": np.random.uniform(1, 20, 900),
    "Age": np.random.randint(22, 60, 900),
    "Gender": np.random.choice(["Male", "Female"], 900),
    "Rating": np.random.choice([1,2,3,4,5], 900, p=[0.05,0.1,0.3,0.35,0.2])
})

# Global Seaborn theme
sns.set_theme(style="whitegrid", font_scale=1.1)

🔍 Kya Hai: Seaborn ka histplot() matplotlib ke hist() ka upgraded version hai — KDE overlay, hue-based split, multiple stat modes, aur DataFrame direct support ke saath. Distribution analysis ka primary tool.

💻 Basic Version:

fig, ax = plt.subplots(figsize=(10, 6))
sns.histplot(data=df, x="Salary", ax=ax)
ax.set_title("Salary Distribution")
plt.savefig("hist_basic.png", dpi=150, bbox_inches="tight")
plt.show()

💻 Professional Styled Version:

fig, ax = plt.subplots(figsize=(12, 6))

sns.histplot(data=df, x="Salary",
    bins=30,                       # number of bins
    kde=True,                      # overlay KDE curve
    color="#3498db",               # bar color
    edgecolor="white",             # bar border
    linewidth=0.8,                 # border thickness
    alpha=0.7,                     # transparency
    stat="count",                  # "count","density","probability","percent","frequency"
    element="bars",                # "bars","step","poly"
    fill=True,                     # fill bars with color
    kde_kws={                       # KDE curve styling
        "color": "#e74c3c",
        "linewidth": 2.5,
        "label": "KDE Curve"
    },
    ax=ax
)

# Mean & Median lines
ax.axvline(df["Salary"].mean(), color="red", linewidth=2, linestyle="--",
           label=f"Mean: ₹{df['Salary'].mean():,.0f}")
ax.axvline(df["Salary"].median(), color="green", linewidth=2, linestyle="-.",
           label=f"Median: ₹{df['Salary'].median():,.0f}")

ax.set_title("Salary Distribution with KDE", fontsize=16, fontweight="bold", pad=15)
ax.set_xlabel("Salary (₹)", fontsize=12)
ax.set_ylabel("Count", fontsize=12)
ax.legend(fontsize=11)
sns.despine()
plt.savefig("hist_styled.png", dpi=150, bbox_inches="tight")
plt.show()

💻 Hue Split — Department-wise Distribution:

fig, ax = plt.subplots(figsize=(12, 6))

sns.histplot(data=df, x="Salary",
    hue="Department",              # split by category — separate colors!
    multiple="stack",              # "layer","dodge","stack","fill"
    bins=25,
    palette="Set2",               # color palette
    edgecolor="white",
    alpha=0.8,
    ax=ax
)

ax.set_title("Salary Distribution by Department (Stacked)", fontsize=15, fontweight="bold")
sns.despine()
plt.savefig("hist_hue.png", dpi=150, bbox_inches="tight")
plt.show()

📋 All Parameters:

🔍 Kya Hai: KDE plot histogram ka smooth continuous version hai — bars ki jagah smooth curve se distribution dikhata hai. Multiple distributions overlap karna, bimodal patterns detect karna, aur probability density visualize karna — iske liye KDE best hai.

💻 Basic Version:

fig, ax = plt.subplots(figsize=(10, 6))
sns.kdeplot(data=df, x="Salary", ax=ax)
ax.set_title("Salary Density Curve")
plt.savefig("kde_basic.png", dpi=150, bbox_inches="tight")
plt.show()

💻 Professional — Multiple Groups + Filled:

fig, ax = plt.subplots(figsize=(12, 6))

sns.kdeplot(data=df, x="Salary",
    hue="Department",              # split by department
    fill=True,                     # fill area under curve
    alpha=0.4,                     # fill transparency
    linewidth=2,                   # curve line thickness
    palette="Set2",               # color palette
    bw_adjust=0.8,                # bandwidth (lower=more detail, higher=smoother)
    common_norm=False,             # each group separately normalized
    cut=0,                         # clip curve at data limits (no extending)
    ax=ax
)

ax.set_title("Salary Density by Department", fontsize=16, fontweight="bold")
ax.set_xlabel("Salary (₹)", fontsize=12)
ax.set_ylabel("Density", fontsize=12)
sns.despine()
plt.savefig("kde_groups.png", dpi=150, bbox_inches="tight")
plt.show()

💻 2D KDE — Bivariate Density (Contour):

fig, ax = plt.subplots(figsize=(10, 7))

sns.kdeplot(data=df, x="Experience", y="Salary",
    fill=True,                     # filled contours
    cmap="YlOrRd",                # colormap for density
    levels=15,                    # number of contour levels
    thresh=0.05,                  # minimum density threshold
    alpha=0.8,
    ax=ax
)

# Overlay scatter for actual data points
ax.scatter(df["Experience"], df["Salary"], color="black", alpha=0.1, s=5, zorder=2)

ax.set_title("Experience vs Salary — 2D Density", fontsize=15, fontweight="bold")
ax.set_xlabel("Experience (Years)", fontsize=12)
ax.set_ylabel("Salary (₹)", fontsize=12)
plt.savefig("kde_2d.png", dpi=150, bbox_inches="tight")
plt.show()

📋 KDE-Specific Parameters:

🔍 Kya Hai: Rugplot axis par chhote vertical tick marks lagata hai — har mark ek actual data point represent karta hai. Akela rarely use hota hai, lekin histogram ya KDE ke saath combine karne se data density clearly dikhti hai.

💻 Combined — KDE + Rug:

fig, ax = plt.subplots(figsize=(12, 6))

# KDE curve
sns.kdeplot(data=df, x="Salary", fill=True, color="#3498db", alpha=0.3, linewidth=2, ax=ax)

# Rug marks on x-axis
sns.rugplot(data=df, x="Salary",
    height=0.05,                  # tick mark height (0-1 fraction of plot)
    color="#e74c3c",               # tick color
    alpha=0.3,                     # transparency (important for overlapping!)
    linewidth=1,                   # tick thickness
    ax=ax
)

ax.set_title("Salary Distribution — KDE + Rug", fontsize=15, fontweight="bold")
ax.set_xlabel("Salary (₹)", fontsize=12)
sns.despine()
plt.savefig("kde_rug.png", dpi=150, bbox_inches="tight")
plt.show()

💻 Hue-Based Rug with Histogram:

fig, ax = plt.subplots(figsize=(12, 6))

sns.histplot(data=df, x="Salary", hue="Gender", kde=True,
    multiple="layer", alpha=0.5, palette=["#3498db","#e74c3c"], ax=ax)

sns.rugplot(data=df, x="Salary", hue="Gender",
    height=0.03,
    palette=["#3498db","#e74c3c"],
    alpha=0.2,
    ax=ax
)

ax.set_title("Salary by Gender — Hist + KDE + Rug", fontsize=15, fontweight="bold")
sns.despine()
plt.savefig("hist_kde_rug.png", dpi=150, bbox_inches="tight")
plt.show()

📋 Rug Parameters:

🔍 Kya Hai: ECDF plot data ki cumulative probability dikhata hai — x-axis par value aur y-axis par "kitne percent data is value se neeche hai". Percentile analysis, threshold decisions, aur distribution comparison ke liye histogram se zyada informative hai.

💻 Basic Version:

fig, ax = plt.subplots(figsize=(10, 6))

sns.ecdfplot(data=df, x="Salary",
    stat="proportion",            # "proportion" (0-1) or "count" (raw)
    complementary=False,          # True = survival function (1 - CDF)
    linewidth=2.5,
    color="#3498db",
    ax=ax
)

ax.set_title("Salary ECDF — Cumulative Distribution", fontsize=15, fontweight="bold")
ax.set_xlabel("Salary (₹)")
ax.set_ylabel("Proportion")
sns.despine()
plt.savefig("ecdf_basic.png", dpi=150, bbox_inches="tight")
plt.show()

💻 Professional — Multi-Group + Percentile Lines:

fig, ax = plt.subplots(figsize=(12, 7))

sns.ecdfplot(data=df, x="Salary",
    hue="Department",
    linewidth=2.5,
    palette="Set2",
    ax=ax
)

# Percentile reference lines
for pct, style in [(0.25, ":"), (0.50, "--"), (0.75, "-.")]:
    val = df["Salary"].quantile(pct)
    ax.axhline(y=pct, color="gray", linestyle=style, alpha=0.5)
    ax.axvline(x=val, color="gray", linestyle=style, alpha=0.5)
    ax.text(val + 1000, pct + 0.02, f"P{int(pct*100)}: ₹{val:,.0f}",
            fontsize=9, color="#7f8c8d")

ax.set_title("Salary ECDF by Department + Percentiles", fontsize=15, fontweight="bold")
ax.set_xlabel("Salary (₹)", fontsize=12)
ax.set_ylabel("Cumulative Proportion", fontsize=12)
sns.despine()
plt.savefig("ecdf_percentiles.png", dpi=150, bbox_inches="tight")
plt.show()

🔍 Kya Hai: displot() Seaborn ka figure-level function hai jo histplot, kdeplot, aur ecdfplot teeno ko kind= parameter se control karta hai. Iska sabse powerful feature hai col= aur row= parameters jo automatic multi-panel faceted charts banate hain.

💻 Faceted Histogram — Department-wise Panels:

# Faceted histogram — separate panel for each department
g = sns.displot(data=df, x="Salary",
    col="Department",              # separate column for each department
    kind="hist",                  # "hist", "kde", "ecdf"
    kde=True,                      # overlay KDE on histogram
    bins=20,
    col_wrap=2,                   # 2 charts per row, wrap to next line
    height=4,                     # each panel height (inches)
    aspect=1.5,                   # width = height × aspect
    color="#3498db",
    edgecolor="white"
)

g.fig.suptitle("Salary Distribution — Each Department", fontsize=16, fontweight="bold", y=1.03)
plt.savefig("displot_faceted.png", dpi=150, bbox_inches="tight")
plt.show()

💻 Faceted KDE — Row + Col Grid:

# 2D facet grid — Department (columns) × Gender (rows)
g = sns.displot(data=df, x="Salary",
    col="Department",              # columns
    row="Gender",                  # rows
    kind="kde",                   # KDE curves
    fill=True,
    height=3.5,
    aspect=1.3,
    palette="Set2",
    alpha=0.6
)

g.fig.suptitle("Salary KDE — Department × Gender Grid", fontsize=16, fontweight="bold", y=1.03)
plt.savefig("displot_grid.png", dpi=150, bbox_inches="tight")
plt.show()

💻 ECDF Mode:

# ECDF mode with hue
g = sns.displot(data=df, x="Salary",
    hue="Gender",
    kind="ecdf",                  # ECDF mode
    height=5,
    aspect=1.8,
    palette=["#3498db", "#e74c3c"],
    linewidth=2.5
)

g.fig.suptitle("Salary ECDF by Gender", fontsize=15, fontweight="bold", y=1.03)
plt.savefig("displot_ecdf.png", dpi=150, bbox_inches="tight")
plt.show()

📋 displot() Parameters:

🔍 Kya Hai: Production-ready distribution analysis — 4 panels mein complete picture: Histogram+KDE, KDE groups, ECDF, aur Box+Violin combined. Copy-paste ready template jo EDA mein directly use ho.

💻 Complete 4-Panel Distribution Dashboard:

fig, axes = plt.subplots(2, 2, figsize=(16, 12))
fig.patch.set_facecolor("#f8fafc")

# Panel 1: Histogram + KDE + Rug
sns.histplot(data=df, x="Salary", bins=30, kde=True, color="#3498db",
    edgecolor="white", alpha=0.7, ax=axes[0][0])
sns.rugplot(data=df, x="Salary", height=0.03, color="#e74c3c", alpha=0.2, ax=axes[0][0])
axes[0][0].axvline(df["Salary"].mean(), color="red", linestyle="--", linewidth=1.5,
    label=f"Mean: ₹{df['Salary'].mean():,.0f}")
axes[0][0].set_title("📊 Histogram + KDE + Rug", fontweight="bold")
axes[0][0].legend(fontsize=9)

# Panel 2: KDE by Department
sns.kdeplot(data=df, x="Salary", hue="Department", fill=True,
    alpha=0.4, linewidth=2, palette="Set2", common_norm=False, ax=axes[0][1])
axes[0][1].set_title("📈 KDE by Department", fontweight="bold")

# Panel 3: ECDF with Percentiles
sns.ecdfplot(data=df, x="Salary", hue="Gender",
    linewidth=2.5, palette=["#3498db","#e74c3c"], ax=axes[1][0])
axes[1][0].axhline(y=0.5, color="gray", linestyle="--", alpha=0.5)
axes[1][0].set_title("📉 ECDF by Gender", fontweight="bold")

# Panel 4: Box + Strip (distribution + individual points)
sns.boxplot(data=df, x="Department", y="Salary", palette="Set2",
    width=0.5, fliersize=3, ax=axes[1][1])
sns.stripplot(data=df, x="Department", y="Salary", color="black",
    alpha=0.1, size=2, jitter=True, ax=axes[1][1])
axes[1][1].set_title("📦 Box + Strip by Department", fontweight="bold")

# Clean up all panels
for ax in axes.flat:
    sns.despine(ax=ax)

fig.suptitle("🏢 Salary Distribution — Complete EDA Dashboard",
    fontsize=18, fontweight="bold", y=1.02)
plt.tight_layout()
plt.savefig("distribution_dashboard.png", dpi=150, bbox_inches="tight",
    facecolor="#f8fafc")
plt.show()

💻 Quick Stats Print (Pair with Dashboard):

# Quick distribution stats — print alongside dashboard
col = "Salary"
print(f"📊 Distribution Stats for '{col}':")
print(f"   Mean:     ₹{df[col].mean():,.0f}")
print(f"   Median:   ₹{df[col].median():,.0f}")
print(f"   Std Dev:  ₹{df[col].std():,.0f}")
print(f"   Skewness: {df[col].skew():.3f}")
print(f"   Kurtosis: {df[col].kurtosis():.3f}")
print(f"   Range:    ₹{df[col].min():,.0f} — ₹{df[col].max():,.0f}")
print(f"   IQR:      ₹{df[col].quantile(0.75) - df[col].quantile(0.25):,.0f}")

Next Post: Part 2

Next part mein hum cover karenge: Categorical Charts — barplot(), countplot(), boxplot(), violinplot(), swarmplot(), stripplot() aur catplot() ke saath professional categorical data visualization.

Happy Visualizing! 📊🚀