Scatter Plot in Plotly

Last Updated: 08 Nov 2025

---

Scatter plots show individual data points as markers — perfect for correlation, clustering, outliers, or 2D relationships.

Hinglish Tip: Scatter plot = "har point ko alag dikhana" — jaise stars in the sky. No line, just dots with meaning.

---

Basic Scatter Plot

import plotly.express as px

# Sample data
x = [1, 2, 3, 4, 5]
y = [10, 15, 13, 18, 22]

fig = px.scatter(x=x, y=y, title="Basic Scatter")
fig.show()

---

Key Parameters in px.scatter()

Parameter	Values / Shorthand	Purpose
data_frame	df (Pandas DataFrame)	Use real dataset
x, y	Column names or lists	X and Y axis data
color	Column name or list	Color by category
size	Column name	Bubble size
hover_name	Column	Label on hover
hover_data	List of columns	Extra info on hover
symbol	Column or marker type	Different shapes
title	String	Plot title
labels	Dict: 'col': 'New Name'	Rename axes
width, height	Pixels (e.g. 800)	Figure size

---

Full Featured Example

import plotly.express as px
import pandas as pd

# Sample dataset
df = pd.DataFrame({
    'City': ['Delhi', 'Mumbai', 'Bangalore', 'Chennai', 'Kolkata'],
    'Temperature': [32, 30, 28, 35, 29],
    'Humidity': [60, 80, 70, 75, 85],
    'Pollution': [120, 95, 80, 110, 130],
    'Region': ['North', 'West', 'South', 'South', 'East']
})

fig = px.scatter(
    df,
    x='Temperature',
    y='Humidity',
    size='Pollution',
    color='Region',
    hover_name='City',
    hover_data={'Pollution': True, 'Temperature': False},
    symbol='Region',
    title='City Climate Analysis: Temp vs Humidity',
    labels={
        'Temperature': 'Temp (°C)',
        'Humidity': 'Humidity (%)',
        'Pollution': 'AQI Level'
    },
    width=900,
    height=600
)

fig.update_layout(
    legend_title="Region",
    font=dict(size=12)
)

fig.show()

Hinglish Tip: size='Pollution' → bada bubble = zyada pollution — ek nazar mein samajh aayega!

---

Add Trendline (Regression)

fig = px.scatter(df, x='Temperature', y='Humidity', trendline="ols", title="With Trendline")
fig.show()

Options: "ols", "lowess", "rolling", "expanding"

---

Common Mistakes to Avoid

Mistake	Fix
No hover info	Use hover_name + hover_data
Too many points → slow	Use render_mode='webgl' for 100k+ points
Default colors hard to read	Use color_discrete_sequence=px.colors.qualitative.Bold
Can't export	Install kaleido: pip install kaleido
Small plot	Set width=900, height=600

---