Lesson 4 — Intermediate

Matplotlib se Graphs — Line, Bar aur Histogram

Numbers table me dikhte hain, trend graph me. Is lesson me hum sales data se teen real graphs banayenge — line (trend), bar (category comparison) aur histogram (distribution) — saare dark-theme visuals jo report me seedhe lag sakte hain.

⏱ 30 min• Intermediate• Jupyter / Colab

Figure–Axes ka anatomy

WHAT

Matplotlib me Figure = poora canvas, Axes = ek plot area (jisme lines bhari jaati hain). OO style (fig, ax = plt.subplots()) production me preferred hai kyunki ek figure me multiple axes handle kar sakte ho.

WHEN

EDA ho, dashboard export ho, ya interview me "graph bana ke dikhao" — Matplotlib har Python data stack ka default renderer hai.

WHERE

Reports, Jupyter notebooks, seaborn (jo andar Matplotlib hi use karta hai), aur model training curves (loss vs epoch). Ek baar anatomy samajh lo, plotly/ggplot bhi jaldi seekh jaoge.

Line chart — monthly sales trend

import matplotlib.pyplot as plt
import numpy as np

months = ["Jan","Feb","Mar","Apr","May","Jun",
          "Jul","Aug","Sep","Oct","Nov","Dec"]
sales  = [42, 38, 55, 61, 58, 72, 69, 81, 77, 90, 95, 110]

fig, ax = plt.subplots(figsize=(10, 5), facecolor="#0d0d1c")
ax.set_facecolor("#0d0d1c")

ax.plot(months, sales, marker="o", linewidth=2.5,
        color="#f0a500", label="Monthly Sales (K)")

ax.set_title("Monthly Sales Trend 2026",
             color="white", fontsize=16, pad=12)
ax.set_xlabel("Month", color="#8a8aa3")
ax.set_ylabel("Sales (₹ 000)", color="#8a8aa3")
ax.tick_params(colors="#8a8aa3")
ax.legend(facecolor="#141428", labelcolor="white")
ax.grid(True, alpha=0.15)

plt.tight_layout()
plt.savefig("sales_line.png", dpi=150, bbox_inches="tight")
plt.show()

Monthly sales line chart with gold trend line on dark background — Line chart — 12 mahine ka sales trend (Matplotlib, dark theme)

Bar chart — category comparison

categories = ["Electronics", "Fashion", "Grocery", "Home", "Books"]
revenue    = [420, 310, 520, 260, 140]

fig, ax = plt.subplots(figsize=(9, 5), facecolor="#0d0d1c")
ax.set_facecolor("#0d0d1c")

bars = ax.bar(categories, revenue,
              color=["#f0a500", "#8b5cf6", "#2dd4bf",
                     "#f0a500", "#8b5cf6"],
              edgecolor="rgba(255,255,255,.15)")

# Har bar pe value label
for b in bars:
    ax.text(b.get_x() + b.get_width()/2, b.get_height() + 8,
            f"₹{b.get_height()}K", ha="center",
            color="white", fontsize=10, fontweight="bold")

ax.set_title("Category-wise Revenue", color="white", fontsize=16)
ax.tick_params(colors="#8a8aa3")
ax.grid(axis="y", alpha=0.12)
plt.tight_layout(); plt.show()

Category-wise revenue bar chart in violet gold teal colors on dark background — Bar chart — category-wise revenue comparison (labels ke saath)

Histogram — distribution samajho

np.random.seed(42)
order_values = np.random.lognormal(mean=5.6, sigma=0.6, size=2000)

fig, ax = plt.subplots(figsize=(9, 5), facecolor="#0d0d1c")
ax.set_facecolor("#0d0d1c")

ax.hist(order_values, bins=40, color="#2dd4bf",
        edgecolor="#0d0d1c", alpha=0.9)
ax.axvline(np.mean(order_values), color="#f0a500",
           linestyle="--", linewidth=2, label=f"Mean ₹{np.mean(order_values):.0f}")

ax.set_title("Order Value Distribution", color="white", fontsize=16)
ax.set_xlabel("Order Value (₹)", color="#8a8aa3")
ax.set_ylabel("Count", color="#8a8aa3")
ax.tick_params(colors="#8a8aa3")
ax.legend(facecolor="#141428", labelcolor="white")
plt.tight_layout(); plt.show()

Order value histogram with teal bars and gold mean line on dark background — Histogram — order values ki distribution + mean line

Chart choice guide

Trend over time → line chart (time axis hamesha x pe).
Category compare → bar chart (horizontal jab labels lambe hon).
Distribution → histogram / KDE.
Relationship do numeric → scatter (agla lesson).
Single big number → bar ya KPI text — pie avoid karo (5 se zyada slice kabhi nahi).

Report tip: Figure hamesha savefig(..., dpi=150, bbox_inches="tight") se save karo — slide me extra white margin nahi aata. Title + axis label + legend teenon zaroori hain; "graph kya dikha raha hai" bina label ke interviewer khud guess nahi karega.

Exercise

Question: Matplotlib me poore canvas (plot areas ke container) ka kya naam hai? (6 letters)

Question: Distribution dikhane wala 4-letter Matplotlib function (bins ke saath)?

Common mistakes

Titles ke bina graphs: Sahiya title/label/legend ke bina graph ka meaning khatam.
Default rcParams: White bg dark report me chipak jaata hai — facecolor set karo ya style sheet use karo.
plt.show() se pehle tight_layout miss: Labels kat jaate hain, clipped text aata hai.
Bar pe 10 categories: 10+ bars me comparison mushkil — top-5 + "Others" karo.

Interview question: "Line aur bar chart me farak kab use karein?" — Line = continuous time-series (trend); Bar = discrete categories (comparison). Time axis pe bar ghalat hai kyunki visual continuity toot jaati hai.

Lesson complete?

Lesson ho gaya? Ab next pe chalte hain — Seaborn Statistical Plots.