Fast Bar Charts¶

The fast_bar() method creates bar-like visualizations using stacked areas. This renders much faster than actual bar charts for large datasets because it uses a single polygon per trace instead of individual rectangles.

In [1]:

import numpy as np
import xarray as xr

from xarray_plotly import config, xpx

config.notebook()

import numpy as np import xarray as xr from xarray_plotly import config, xpx config.notebook()

Basic Example¶

In [2]:

# Quarterly revenue data by product and region
np.random.seed(42)
da = xr.DataArray(
    np.random.rand(4, 3, 2) * 100 + 50,
    dims=["quarter", "product", "region"],
    coords={
        "quarter": ["Q1", "Q2", "Q3", "Q4"],
        "product": ["Widgets", "Gadgets", "Gizmos"],
        "region": ["North", "South"],
    },
    name="revenue",
)

xpx(da).fast_bar()

# Quarterly revenue data by product and region np.random.seed(42) da = xr.DataArray( np.random.rand(4, 3, 2) * 100 + 50, dims=["quarter", "product", "region"], coords={ "quarter": ["Q1", "Q2", "Q3", "Q4"], "product": ["Widgets", "Gadgets", "Gizmos"], "region": ["North", "South"], }, name="revenue", ) xpx(da).fast_bar()

In [3]:

# Comparison with regular bar()
xpx(da).bar()

# Comparison with regular bar() xpx(da).bar()

With Faceting¶

In [4]:

xpx(da).fast_bar(facet_col="region")

xpx(da).fast_bar(facet_col="region")

With Animation¶

In [5]:

# Multi-year data for animation
np.random.seed(123)
da_anim = xr.DataArray(
    np.random.rand(4, 3, 5) * 100 + 20,
    dims=["quarter", "product", "year"],
    coords={
        "quarter": ["Q1", "Q2", "Q3", "Q4"],
        "product": ["Widgets", "Gadgets", "Gizmos"],
        "year": [2020, 2021, 2022, 2023, 2024],
    },
    name="revenue",
)

xpx(da_anim).fast_bar(animation_frame="year")

# Multi-year data for animation np.random.seed(123) da_anim = xr.DataArray( np.random.rand(4, 3, 5) * 100 + 20, dims=["quarter", "product", "year"], coords={ "quarter": ["Q1", "Q2", "Q3", "Q4"], "product": ["Widgets", "Gadgets", "Gizmos"], "year": [2020, 2021, 2022, 2023, 2024], }, name="revenue", ) xpx(da_anim).fast_bar(animation_frame="year")

Faceting + Animation¶

In [6]:

# 4D data: quarter x product x region x year
np.random.seed(456)
da_4d = xr.DataArray(
    np.random.rand(4, 3, 2, 4) * 80 + 30,
    dims=["quarter", "product", "region", "year"],
    coords={
        "quarter": ["Q1", "Q2", "Q3", "Q4"],
        "product": ["Widgets", "Gadgets", "Gizmos"],
        "region": ["North", "South"],
        "year": [2021, 2022, 2023, 2024],
    },
    name="revenue",
)

xpx(da_4d).fast_bar(facet_col="region", animation_frame="year")

# 4D data: quarter x product x region x year np.random.seed(456) da_4d = xr.DataArray( np.random.rand(4, 3, 2, 4) * 80 + 30, dims=["quarter", "product", "region", "year"], coords={ "quarter": ["Q1", "Q2", "Q3", "Q4"], "product": ["Widgets", "Gadgets", "Gizmos"], "region": ["North", "South"], "year": [2021, 2022, 2023, 2024], }, name="revenue", ) xpx(da_4d).fast_bar(facet_col="region", animation_frame="year")

Positive and Negative Values¶

fast_bar() classifies each trace by its values:

• Purely positive → stacks upward
• Purely negative → stacks downward
• Mixed signs → warning + dashed line (use bar() instead)

In [7]:

# Profit (positive) and Loss (negative) - stacks correctly
np.random.seed(789)
da_split = xr.DataArray(
    np.column_stack(
        [
            np.random.rand(6) * 80 + 20,  # Revenue: positive
            -np.random.rand(6) * 50 - 10,  # Costs: negative
        ]
    ),
    dims=["month", "category"],
    coords={
        "month": ["Jan", "Feb", "Mar", "Apr", "May", "Jun"],
        "category": ["Revenue", "Costs"],
    },
    name="financials",
)

xpx(da_split).fast_bar()

# Profit (positive) and Loss (negative) - stacks correctly np.random.seed(789) da_split = xr.DataArray( np.column_stack( [ np.random.rand(6) * 80 + 20, # Revenue: positive -np.random.rand(6) * 50 - 10, # Costs: negative ] ), dims=["month", "category"], coords={ "month": ["Jan", "Feb", "Mar", "Apr", "May", "Jun"], "category": ["Revenue", "Costs"], }, name="financials", ) xpx(da_split).fast_bar()

In [8]:

# With animation - sign classification is consistent across frames
np.random.seed(321)
da_split_anim = xr.DataArray(
    np.stack(
        [
            np.column_stack([np.random.rand(6) * 80 + 20, -np.random.rand(6) * 50 - 10])
            for _ in range(4)
        ],
        axis=-1,
    ),
    dims=["month", "category", "year"],
    coords={
        "month": ["Jan", "Feb", "Mar", "Apr", "May", "Jun"],
        "category": ["Revenue", "Costs"],
        "year": [2021, 2022, 2023, 2024],
    },
    name="financials",
)

xpx(da_split_anim).fast_bar(animation_frame="year")

# With animation - sign classification is consistent across frames np.random.seed(321) da_split_anim = xr.DataArray( np.stack( [ np.column_stack([np.random.rand(6) * 80 + 20, -np.random.rand(6) * 50 - 10]) for _ in range(4) ], axis=-1, ), dims=["month", "category", "year"], coords={ "month": ["Jan", "Feb", "Mar", "Apr", "May", "Jun"], "category": ["Revenue", "Costs"], "year": [2021, 2022, 2023, 2024], }, name="financials", ) xpx(da_split_anim).fast_bar(animation_frame="year")

Mixed Sign Warning¶

When a trace has both positive and negative values, fast_bar() shows a warning and displays it as a dashed line:

In [9]:

# Both columns have mixed signs - triggers warning
da_mixed = xr.DataArray(
    np.array(
        [
            [50, -30],
            [-40, 60],
            [30, -50],
            [-20, 40],
        ]
    ),
    dims=["month", "category"],
    coords={
        "month": ["Jan", "Feb", "Mar", "Apr"],
        "category": ["A", "B"],
    },
)

# This will show a warning
xpx(da_mixed).fast_bar()

# Both columns have mixed signs - triggers warning da_mixed = xr.DataArray( np.array( [ [50, -30], [-40, 60], [30, -50], [-20, 40], ] ), dims=["month", "category"], coords={ "month": ["Jan", "Feb", "Mar", "Apr"], "category": ["A", "B"], }, ) # This will show a warning xpx(da_mixed).fast_bar()

/home/runner/work/xarray_plotly/xarray_plotly/xarray_plotly/accessor.py:199: UserWarning:

fast_bar: traces ['A', 'B'] have mixed positive/negative values and cannot be stacked. They are shown as dashed lines. Consider using bar() for proper stacking of mixed data.

In [10]:

# For mixed data, use bar() instead
xpx(da_mixed).bar()

# For mixed data, use bar() instead xpx(da_mixed).bar()

When to Use¶

Method	Use when...
fast_bar()	Large datasets, animations, performance matters, data is same-sign per trace
bar()	Need grouped bars, pattern fills, or have mixed +/- values per trace
area()	Want smooth continuous fills