Custom Plot

Create fully customized visualizations by writing your own Python and Plotly code. This gives you complete control over your plots with access to the full Plotly library and your data.

When to Use

Need a specific visualization not available in standard plots
Want to combine multiple plot types in one figure
Require advanced customization beyond standard options
Have specific design requirements
Need to implement complex data transformations
Want to create specialized domain-specific visualizations

Available Variables

Your code has access to these pre-loaded libraries and variables:

Data

df - Your Pandas DataFrame containing the data
pd - Pandas library for data manipulation
np - NumPy library for numerical operations

Plotting Libraries

px - Plotly Express for quick, high-level plots
go - Plotly Graph Objects for detailed, low-level control
colors - Pre-defined color palette from your theme

Requirements

Your code must:

Create a Plotly figure object
Assign it to the variable fig
Be valid Python code

Example:

fig = px.scatter(df, x='column1', y='column2', title='My Custom Plot')

Basic Examples

Simple Scatter Plot

import plotly.express as px

fig = px.scatter(
    df,
    x='age',
    y='salary',
    title='Age vs Salary',
    labels={'age': 'Age (years)', 'salary': 'Annual Salary'}
)

Bar Chart with Aggregation

# Group and aggregate data
grouped = df.groupby('category')['value'].sum().reset_index()

fig = px.bar(
    grouped,
    x='category',
    y='value',
    title='Total Value by Category'
)

Line Chart with Multiple Lines

fig = px.line(
    df,
    x='date',
    y=['metric1', 'metric2', 'metric3'],
    title='Multiple Metrics Over Time'
)

Custom Colors

fig = px.scatter(
    df,
    x='x_col',
    y='y_col',
    color='category',
    color_discrete_sequence=colors  # Use theme colors
)

Advanced Examples

Subplots with Multiple Chart Types

from plotly.subplots import make_subplots
import plotly.graph_objects as go

# Create subplots
fig = make_subplots(
    rows=2, cols=1,
    subplot_titles=('Sales Over Time', 'Profit Margin')
)

# Add line chart
fig.add_trace(
    go.Scatter(x=df['date'], y=df['sales'], name='Sales'),
    row=1, col=1
)

# Add bar chart
fig.add_trace(
    go.Bar(x=df['date'], y=df['profit_margin'], name='Profit Margin'),
    row=2, col=1
)

fig.update_layout(height=600, title_text="Sales Dashboard")

Heatmap with Custom Styling

# Create pivot table
pivot = df.pivot_table(
    values='value',
    index='row_category',
    columns='col_category',
    aggfunc='mean'
)

fig = go.Figure(data=go.Heatmap(
    z=pivot.values,
    x=pivot.columns,
    y=pivot.index,
    colorscale='Viridis',
    text=pivot.values,
    texttemplate='%{text:.2f}',
    textfont={"size": 10}
))

fig.update_layout(
    title='Correlation Heatmap',
    xaxis_title='Categories',
    yaxis_title='Groups'
)

3D Scatter Plot

fig = go.Figure(data=[go.Scatter3d(
    x=df['x'],
    y=df['y'],
    z=df['z'],
    mode='markers',
    marker=dict(
        size=5,
        color=df['value'],
        colorscale='Viridis',
        showscale=True
    )
)])

fig.update_layout(
    title='3D Visualization',
    scene=dict(
        xaxis_title='X Axis',
        yaxis_title='Y Axis',
        zaxis_title='Z Axis'
    )
)

Animated Plot

fig = px.scatter(
    df,
    x='gdp_per_capita',
    y='life_expectancy',
    animation_frame='year',
    animation_group='country',
    size='population',
    color='continent',
    hover_name='country',
    log_x=True,
    range_x=[100, 100000],
    range_y=[25, 90],
    title='Gapminder Animation'
)

Box Plot with Statistical Annotations

fig = go.Figure()

for category in df['category'].unique():
    data = df[df['category'] == category]['value']
    fig.add_trace(go.Box(
        y=data,
        name=category,
        boxmean='sd'  # Show mean and standard deviation
    ))

fig.update_layout(
    title='Distribution by Category',
    yaxis_title='Value',
    showlegend=True
)

Data Manipulation Examples

Filtering and Transforming

# Filter data
filtered_df = df[df['value'] &gt; 100]

# Create new columns
filtered_df['log_value'] = np.log(filtered_df['value'])

# Create plot
fig = px.histogram(
    filtered_df,
    x='log_value',
    title='Distribution of Log Values (filtered)'
)

Rolling Averages

# Calculate 7-day rolling average
df_sorted = df.sort_values('date')
df_sorted['rolling_avg'] = df_sorted['value'].rolling(window=7).mean()

fig = go.Figure()
fig.add_trace(go.Scatter(
    x=df_sorted['date'],
    y=df_sorted['value'],
    name='Actual',
    mode='lines',
    line=dict(color='lightgray')
))
fig.add_trace(go.Scatter(
    x=df_sorted['date'],
    y=df_sorted['rolling_avg'],
    name='7-Day Average',
    mode='lines',
    line=dict(color='blue', width=2)
))

fig.update_layout(title='Time Series with Rolling Average')

Pivot and Reshape

# Reshape data
pivot_df = df.pivot_table(
    values='sales',
    index='month',
    columns='product',
    aggfunc='sum'
)

fig = px.imshow(
    pivot_df,
    labels=dict(x="Product", y="Month", color="Sales"),
    title='Sales Heatmap'
)

Styling and Layout

Custom Theme

fig = px.scatter(df, x='x', y='y', color='category')

fig.update_layout(
    template='plotly_dark',  # or 'plotly', 'seaborn', 'ggplot2', etc.
    font=dict(family='Arial', size=12),
    title=dict(text='Custom Styled Plot', x=0.5, xanchor='center'),
    plot_bgcolor='#1e1e1e',
    paper_bgcolor='#2e2e2e',
    margin=dict(l=50, r=50, t=80, b=50)
)

Custom Hover Templates

fig = px.scatter(df, x='x', y='y', color='category')

fig.update_traces(
    hovertemplate='<b>%{fullData.name}</b><br>' +
                  'X: %{x:.2f}<br>' +
                  'Y: %{y:.2f}<br>' +
                  '<extra></extra>'
)

Annotations and Shapes

fig = px.line(df, x='date', y='value')

# Add annotation
fig.add_annotation(
    x='2023-06-15',
    y=1000,
    text='Peak Sales',
    showarrow=True,
    arrowhead=2
)

# Add horizontal line
fig.add_hline(
    y=750,
    line_dash='dash',
    line_color='red',
    annotation_text='Target'
)

# Add shaded region
fig.add_vrect(
    x0='2023-01-01',
    x1='2023-03-31',
    fillcolor='green',
    opacity=0.2,
    annotation_text='Q1'
)

Common Patterns

Conditional Coloring

# Add color column based on condition
df['color'] = df['value'].apply(
    lambda x: 'green' if x &gt; 0 else 'red'
)

fig = px.bar(df, x='category', y='value', color='color',
             color_discrete_map={'green': '#00cc00', 'red': '#cc0000'})
fig.update_layout(showlegend=False)

Multiple Y-Axes

fig = make_subplots(specs=[[{"secondary_y": True}]])

fig.add_trace(
    go.Scatter(x=df['date'], y=df['sales'], name='Sales'),
    secondary_y=False
)

fig.add_trace(
    go.Scatter(x=df['date'], y=df['profit_margin'], name='Profit %'),
    secondary_y=True
)

fig.update_xaxes(title_text='Date')
fig.update_yaxes(title_text='Sales ($)', secondary_y=False)
fig.update_yaxes(title_text='Profit Margin (%)', secondary_y=True)

Faceted / Small Multiples

fig = px.scatter(
    df,
    x='x',
    y='y',
    facet_col='category',  # Create columns
    facet_row='region',     # Create rows
    title='Faceted Scatter Plot'
)

fig.update_xaxes(matches=None)  # Independent x-axes
fig.update_yaxes(matches=None)  # Independent y-axes

Error Handling

Safe Data Access

# Check if columns exist
required_cols = ['x', 'y']
if all(col in df.columns for col in required_cols):
    fig = px.scatter(df, x='x', y='y')
else:
    # Create error message plot
    fig = go.Figure()
    fig.add_annotation(
        text=f"Required columns not found: {required_cols}",
        xref="paper", yref="paper",
        x=0.5, y=0.5, showarrow=False,
        font=dict(size=16, color="red")
    )

Handle Missing Data

# Remove or fill missing values
df_clean = df.dropna(subset=['x', 'y'])
# Or fill with mean
# df['y'] = df['y'].fillna(df['y'].mean())

fig = px.scatter(df_clean, x='x', y='y')

Performance Tips

Filter data first - Work with smaller datasets when possible

df_sample = df.sample(n=1000)  # Use sample for large datasets
fig = px.scatter(df_sample, x='x', y='y')

Use appropriate plot types - Scattergl for large scatter plots

fig = go.Figure(data=go.Scattergl(x=df['x'], y=df['y'], mode='markers'))

Aggregate before plotting - Reduce data points

df_agg = df.groupby('date').agg({'value': 'mean'}).reset_index()
fig = px.line(df_agg, x='date', y='value')

Best Practices

Always assign to fig - Your code must create and assign to fig
Test with small data - Verify logic before running on full dataset
Add titles and labels - Make plots self-explanatory
Use meaningful colors - Leverage the colors variable for theme consistency
Handle edge cases - Check for empty data, missing columns, etc.
Comment your code - Explain complex logic
Keep it simple - Start basic, add complexity gradually
Use built-in plots when possible - Custom code is for special cases

Debugging Tips

If your plot doesn't appear:

Check that you assigned to fig (not figure or other name)
Verify column names exist in df
Look for Python syntax errors
Try a simple example first (e.g., fig = px.scatter(df))

Common errors:

KeyError: Column name doesn't exist in DataFrame
TypeError: Wrong data type for plot parameter
ValueError: Data shape or format issue

Quick test:

# Minimal working example
print(df.columns)  # See available columns
print(df.head())   # Preview data
fig = px.scatter(df, x=df.columns[0], y=df.columns[1])

Resources

Limitations

Pro tier required for custom code execution
Code runs in a sandboxed environment
Execution timeout after 30 seconds
Maximum DataFrame size: 1M rows (sampling recommended for larger datasets)
Cannot import additional libraries (only pre-loaded ones available)
Cannot make external API calls or file system access

Remember: With great power comes great responsibility. Custom plots give you unlimited flexibility, but standard plots are optimized, tested, and often sufficient for most use cases. Use custom plots when you truly need the extra control.