OpenBB-finance/OpenBB · error · ValueError

Error: expected data with numeric values.

Error message

Error: expected data with numeric values.

What it means

In create_line_chart's auto_layout branch, columns are ordered by their max-min spread computed with numeric_only=True. If no column is numeric, the sorted index is empty and the ValueError 'expected data with numeric values' is raised - the multi-axis layout algorithm needs at least one numeric series to assign to y-axes.

Source

Thrown at openbb_platform/obbject_extensions/charting/openbb_charting/charts/generic_charts.py:150

    y2 = y2 if y2 else []
    yaxis_num = 1
    yaxis = f"y{yaxis_num}"
    first_y = y[0]  # type: ignore[index]
    second_y = None
    third_y = None
    add_scatter = False

    # Attempt to layout the chart automatically with multiple y-axis.
    mode = scatter_kwargs.pop("mode", "lines")
    hovertemplate = scatter_kwargs.pop("hovertemplate", None)

    if auto_layout is True:
        # Sort columns by the difference between the max and min values.
        # This is to help determine which columns should share the same y-axis.
        diff = df.max(numeric_only=True) - df.min(numeric_only=True)
        sorted_columns = diff.sort_values(ascending=False).index
        if sorted_columns is None or len(sorted_columns) == 0:
            raise ValueError("Error: expected data with numeric values.")
        df = df[sorted_columns]  # type: ignore

        for i, col in enumerate(df.columns):
            if col in y:  # type: ignore[operator]
                hovertemplate = (
                    hovertemplate
                    if hovertemplate
                    else f"{df[col].name}: %{{y}}<extra></extra>"
                )
                share_yaxis = should_share_axis(df, first_y, col, threshold=2.5)
                if share_yaxis is True:
                    add_scatter = True
                if share_yaxis is False:
                    yaxis_num = 2
                    yaxis = f"y{yaxis_num}"
                    if second_y is None:
                        second_y = col
                        add_scatter = True

View on GitHub (pinned to 3e071fcc2c)

Solutions

1. Coerce columns to numeric: df = df.apply(pd.to_numeric, errors='coerce')
2. Pass only the numeric columns: create_line_chart(data=df[['price','volume']])
3. Disable auto_layout and specify x/y explicitly

Example fix

# before
fig = create_line_chart(data=df, auto_layout=True)  # all-object dtypes

# after
df = df.apply(pd.to_numeric, errors='coerce').dropna(axis=1, how='all')
fig = create_line_chart(data=df, auto_layout=True)

num = df.select_dtypes('number')
assert not num.empty, 'line chart auto_layout needs numeric columns'
df = df[num.columns.union(df.columns[:1])]

def has_numeric_columns(df: pd.DataFrame) -> bool:
    return not df.select_dtypes('number').empty

try:
    fig = create_line_chart(data=df, auto_layout=True)
except ValueError as e:
    if 'numeric values' in str(e):
        fig = create_line_chart(data=df.apply(pd.to_numeric, errors='coerce').dropna(axis=1, how='all'), auto_layout=True)

When it happens

Trigger: Calling create_line_chart(..., auto_layout=True) with a DataFrame whose columns are all strings/dates/objects; mixed frames where numeric columns are stored as object dtype (e.g. strings '1.2'); auto-layout reached through .charting() on results with only categorical fields.

Common situations: Provider data read from CSV with numbers as strings; date-only DataFrames; columns containing None-heavy object arrays that numeric_only drops entirely.

Related errors

• Error: No data to plot.
• Expiration field not found in the data.
• Price field not found in the data.
• Error: Rate column not found in the data.
• Error: Date column not found in the data.

AI-assisted analysis of OpenBB-finance/OpenBB@3e071fcc2c (2026-08-14). Data as JSON: /api/errors/60b1639203b1d0e5. Report an issue: GitHub.