Python Libraries for Data Analysis India 2026
Pandas, NumPy, Matplotlib, Seaborn, Plotly & More — With Code

pandas

import pandas as pd

# Load Indian sales data
df = pd.read_csv('sales_india.csv')

# Clean: remove duplicates, fix amount column
df = df.drop_duplicates()
df['amount'] = df['amount'].str.replace('₹', '').str.replace(',', '').astype(float)

# GroupBy — total sales per city per month
summary = df.groupby(['city', 'month'])['amount'].sum().reset_index()

# Filter — top cities above ₹10L
top_cities = summary[summary['amount'] > 1000000]

# Merge — add region from lookup table
regions = pd.read_csv('city_regions.csv')
result = top_cities.merge(regions, on='city', how='left')
print(result.head())

NumPy

import numpy as np

amounts = np.array([15000, 82000, 3500, 120000, 47000, 9000])

# Statistical summary
print(f"Mean:   ₹{np.mean(amounts):,.0f}")
print(f"Median: ₹{np.median(amounts):,.0f}")
print(f"Std:    ₹{np.std(amounts):,.0f}")
print(f"75th %: ₹{np.percentile(amounts, 75):,.0f}")

# Conditional: flag high-value orders
category = np.where(amounts > 50000, 'High Value', 'Standard')
print(category)
# → ['Standard' 'High Value' 'Standard' 'High Value' 'Standard' 'Standard']

# Clip outliers to ₹1L cap
capped = np.clip(amounts, 0, 100000)

Matplotlib

import matplotlib.pyplot as plt
import pandas as pd

months = ['Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep']
revenue = [42, 58, 51, 67, 73, 61]   # ₹ Lakhs
target  = [50, 55, 60, 65, 70, 75]

fig, ax = plt.subplots(figsize=(9, 4))

ax.bar(months, revenue, color='#1d4ed8', alpha=0.8, label='Actual Revenue')
ax.plot(months, target, color='#FF6B00', marker='o',
        linewidth=2, linestyle='--', label='Target')

ax.set_title('Monthly Revenue vs Target — FY 2026 (₹ Lakhs)', fontsize=13)
ax.set_ylabel('₹ Lakhs')
ax.legend()
ax.grid(axis='y', alpha=0.3)

plt.tight_layout()
plt.savefig('revenue_vs_target.png', dpi=150)
plt.show()

Seaborn

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

df = pd.read_csv('employee_data.csv')
# columns: department, salary, experience_years, attrition

# Correlation heatmap
fig, axes = plt.subplots(1, 2, figsize=(12, 4))

numeric_cols = df[['salary', 'experience_years', 'age', 'performance_score']]
sns.heatmap(numeric_cols.corr(), annot=True, fmt='.2f',
            cmap='Blues', ax=axes[0])
axes[0].set_title('Correlation Matrix')

# Box plot — salary by department
sns.boxplot(data=df, x='department', y='salary',
            palette='Set2', ax=axes[1])
axes[1].set_title('Salary Distribution by Department')
axes[1].tick_params(axis='x', rotation=45)

plt.tight_layout()
plt.savefig('eda_charts.png', dpi=150)

Plotly

import plotly.express as px
import pandas as pd

df = pd.read_csv('india_state_sales.csv')
# columns: state, revenue, category, year

# Interactive bar chart — hover shows exact values
fig = px.bar(df[df['year'] == 2026],
             x='state', y='revenue',
             color='category',
             title='Revenue by State — India 2026',
             labels={'revenue': 'Revenue (₹ Cr)', 'state': 'State'},
             color_discrete_sequence=px.colors.qualitative.Set2)

fig.update_layout(xaxis_tickangle=-45)
fig.write_html('india_revenue.html')  # Share as interactive HTML
fig.show()

# Choropleth map — India state-level data
fig2 = px.choropleth(df[df['year'] == 2026],
                     geojson='india_states.geojson',
                     locations='state',
                     color='revenue',
                     title='Revenue Heatmap — India 2026')
fig2.show()

Scikit-learn

from sklearn.linear_model import LinearRegression
from sklearn.model_selection import train_test_split
from sklearn.metrics import r2_score, mean_absolute_error
from sklearn.preprocessing import StandardScaler
import pandas as pd

df = pd.read_csv('sales_forecast.csv')
# Predict next month revenue from marketing spend, region, product category

X = df[['marketing_spend', 'region_encoded', 'category_encoded']]
y = df['revenue']

X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.2, random_state=42)

scaler = StandardScaler()
X_train_scaled = scaler.fit_transform(X_train)
X_test_scaled = scaler.transform(X_test)

model = LinearRegression()
model.fit(X_train_scaled, y_train)

y_pred = model.predict(X_test_scaled)
print(f"R² Score: {r2_score(y_test, y_pred):.3f}")
print(f"MAE:      ₹{mean_absolute_error(y_test, y_pred):,.0f}")

SQLAlchemy + psycopg2

import pandas as pd
from sqlalchemy import create_engine

# Connect to MySQL database (common in Indian companies)
engine = create_engine(
    'mysql+pymysql://username:password@localhost:3306/sales_db'
)

# Run SQL query → directly into pandas DataFrame
query = """
    SELECT
        city,
        DATE_FORMAT(order_date, '%Y-%m') AS month,
        SUM(amount)                       AS total_revenue,
        COUNT(*)                          AS order_count
    FROM orders
    WHERE order_date >= '2026-04-01'
    GROUP BY city, month
    ORDER BY month, total_revenue DESC
"""

df = pd.read_sql(query, engine)
print(df.head(10))

# Write cleaned data back to a results table
df_cleaned.to_sql('monthly_summary', engine,
                  if_exists='replace', index=False)

Which Python library should a data analyst learn first in India?

Start with pandas. It is the foundation of almost all data analysis in Python — loading CSVs, cleaning data, filtering rows, grouping, merging tables, and exporting results. Every other library builds on top of pandas DataFrames. NumPy comes alongside pandas naturally (pandas uses NumPy internally). Once you can do basic pandas operations confidently, add Matplotlib for plotting, then Seaborn for statistical visualisation. Scikit-learn comes after these once you are comfortable with the data preparation workflow.

Is pandas enough for a data analyst job in India?

Strong pandas skills combined with SQL and a BI tool (Power BI or Tableau) are sufficient for most data analyst roles in India. Many job descriptions list "Python/pandas" as a skill, and the practical assessment typically involves reading a CSV, cleaning it, and producing a grouped summary or a chart — all achievable with pandas and Matplotlib alone. NumPy, Seaborn, and Plotly add value and are worth learning, but they are secondary to pandas proficiency. Scikit-learn is needed for data science and ML roles, not pure data analyst roles.

What is the difference between Matplotlib and Seaborn in Python?

Matplotlib is the foundational plotting library — it gives you complete control over every element of a chart but requires more code for professional-looking results. Seaborn is built on top of Matplotlib and provides higher-level functions for statistical visualisations (distribution plots, heatmaps, pair plots, regression plots) with much less code and better default styling. For data analysis work, use Seaborn for exploratory analysis and statistical plots; use Matplotlib for fine-tuned customisation and when you need full control. Both are worth knowing — Seaborn for speed, Matplotlib for precision.

Should data analysts in India learn Plotly or Matplotlib?

Both serve different purposes. Matplotlib produces static charts (PNG/PDF) suitable for reports and presentations. Plotly produces interactive charts (hover, zoom, filter) suitable for web dashboards and Jupyter notebooks. For most data analyst jobs in India, Matplotlib and Seaborn are sufficient. Plotly is valuable if you build dashboards in Python (using Plotly Dash) or need interactive EDA in notebooks. If your company uses Streamlit for internal tools or dashboards, Plotly is the natural partner. Learn Matplotlib and Seaborn first, then add Plotly if your role involves interactive reporting.