plotly-calheatmap 0.4

Calendar Plot made with Plotly

plotly-calheatmap

A continuation of plotly-calplot by Bruno Rocha Silva, which is no longer actively maintained.

This project picks up where plotly-calplot left off, providing an interactive calendar heatmap built with Plotly — similar to the contribution graphs on GitHub and GitLab profile pages.

Features

• Interactive calendar heatmaps built with Plotly
• Dataset swap — switch between multiple metrics via dropdown without regenerating the graph (datasets parameter)
• Built-in aggregation — pass raw event data with agg="sum"|"mean"|"count"|"max" instead of pre-aggregating
• Logarithmic color scale — log_scale=True applies log(1+x) so extreme values don't wash out the heatmap
• Custom time groupings — grouping="quarter"|"bimester"|"semester" draws separator lines and replaces axis labels
• Vertical orientation — render months as rows with vertical=True
• Hourly heatmap — hourly_calheatmap() for hour × day grids per month
• Skip empty years — skip_empty_years=True excludes years with no data
• Replace NaNs with zeros — replace_nans_with_zeros=True displays empty dates as 0
• Top & bottom border lines — top_bottom_lines=True draws horizontal lines enclosing each month
• Month gap spacing — extra visual separation between months via month_gap
• Multi-year support with independent tick configurations per subplot
• Year navigation buttons (navigation=True)
• Localization support (locale parameter) for month and day names (e.g. pt_BR, es, fr)
• Customizable hovertemplate with friendly {placeholder} syntax and customdata columns
• Smart colorscales — pass a colors list and scale_type ("linear", "quantile", "quantize", "diverging", "categorical") for automatic interval computation
• Cell annotations — annotations=True or annotations_fmt="%{z:.0f}" displays values/labels inside each cell
• Zero-value distinction — zero_color gives 0-value cells a dedicated color while missing data stays transparent
• Missing-data styling — nan_color assigns a dedicated color to NaN/missing cells, distinguishing them from zero-value cells
• Responsive / auto-sizing — all chart types adapt width to the container automatically; height is computed from the data (overridable via total_height and width)
• Fully customizable colorscales (including custom lists)
• Month separator lines, configurable month label placement, and color scale with label/ticks
• Flexible layout options: gap, margin, font_*, paper_bgcolor, plot_bgcolor, etc.

Documentation

For the full API reference (all parameters for calheatmap() and hourly_calheatmap()), see docs/API.md.

Installation

pip install plotly-calheatmap

Quick Start

from plotly_calheatmap import calheatmap

fig = calheatmap(df, x="date", y="value")
fig.show()

Built-in Aggregation

Pass raw (non-aggregated) event data directly — duplicate dates are grouped and aggregated automatically:

from plotly_calheatmap import calheatmap

# df has multiple rows per date (e.g. individual transactions)
fig = calheatmap(df, x="date", y="amount", agg="sum")
fig.show()

Supported functions: "sum", "mean", "count", "max".

Vertical Orientation

fig = calheatmap(df, x="date", y="value", vertical=True, month_gap=1)

Dataset Swap

Switch between multiple metrics on the same graph via a dropdown menu. Each dataset can have its own colorscale, value range, and legend title. Works with both calheatmap() and hourly_calheatmap(), alongside year navigation.

fig = calheatmap(
    data=df,
    x="date",
    y="sales",
    datasets={
        "Sales": {
            "y": "sales",
            "colorscale": "greens",
            "showscale": "Sales ($)",
            "cmap_min": 0,
        },
        "Activity": {
            "y": "activity",
            "colorscale": "blues",
            "showscale": "Activity (hours)",
        },
    },
    navigation=True,
)

Logarithmic Color Scale

When a few extreme values wash out the rest of the heatmap, use log_scale=True to apply log(1+x) to the color mapping. Hover text still shows original values.

fig = calheatmap(df, x="date", y="value", log_scale=True)

Smart Colorscales & Zero Distinction

Instead of manually defining colorscale intervals, pass a list of colors and let the library compute the positions automatically:

# Old way (manual positions)
colorscale=[[0.0, "#161b22"], [0.01, "#0e4429"], [0.25, "#006d32"], [1.0, "#39d353"]]

# New way (automatic)
colors=["#0e4429", "#006d32", "#26a641", "#39d353"],
zero_color="#161b22"  # dedicated color for value 0

Scale types control how colors are distributed:

# linear (default): colors evenly spaced across data range
fig = calheatmap(df, x="date", y="value",
    colors=["#edf8e9", "#bae4b3", "#74c476", "#31a354", "#006d2c"])

# quantile: each color covers equal number of data points (good for skewed data)
fig = calheatmap(df, x="date", y="value",
    colors=["#edf8e9", "#bae4b3", "#74c476", "#006d2c"],
    scale_type="quantile")

# quantize: data range split into equal mathematical intervals
fig = calheatmap(df, x="date", y="value",
    colors=["#edf8e9", "#bae4b3", "#74c476", "#006d2c"],
    scale_type="quantize")

# diverging: two gradients meeting at a pivot point
fig = calheatmap(df, x="date", y="temp",
    colors=["#2166ac", "#f7f7f7", "#b2182b"],
    scale_type="diverging", pivot=20, symmetric=True)

# categorical: user-defined bins with explicit color per range
fig = calheatmap(df, x="date", y="value",
    scale_type="categorical",
    bins=[(0, 0, "gray"), (1, 3, "lightgreen"), (4, float("inf"), "darkgreen")])

Zero-value distinction — give cells with value 0 a dedicated color, separate from missing data (which stays transparent):

fig = calheatmap(df, x="date", y="commits",
    colors=["#0e4429", "#006d32", "#26a641", "#39d353"],
    zero_color="#161b22")

Missing-data styling — give NaN/missing cells a visible color instead of leaving them transparent:

fig = calheatmap(df, x="date", y="commits",
    colors=["#0e4429", "#006d32", "#26a641", "#39d353"],
    zero_color="#161b22",
    nan_color="#0d1117")

All new parameters also work inside the datasets dict for per-metric configuration.

Cell Annotations

Display values or labels inside each cell:

# Show integer values inside cells
fig = calheatmap(df, x="date", y="score",
    annotations=True, annotations_fmt="%{z:.0f}",
    gap=4, total_height=280)

# Simple boolean — shows raw z values
fig = calheatmap(df, x="date", y="score", annotations=True)

Custom Time Groupings

Draw thicker separator lines at quarter, bimester, or semester boundaries and replace axis labels with group names:

fig = calheatmap(df, x="date", y="value", grouping="quarter", month_lines=True)

Hourly Heatmap

from plotly_calheatmap import hourly_calheatmap

fig = hourly_calheatmap(df, x="datetime_col", y="value")
fig.show()

Skip Empty Years & Replace NaNs

fig = calheatmap(df, x="date", y="value", skip_empty_years=True, replace_nans_with_zeros=True)

Border Lines

Fully enclose each month by combining month_lines with top_bottom_lines:

fig = calheatmap(df, x="date", y="value", month_lines=True, top_bottom_lines=True)

Credits

This project is based on the original work by Bruno Rocha Silva — plotly-calplot.