microsoft/qlib · error · ValueError

df is empty.

Error message

df is empty.

What it means

BaseGraph._init_data (qlib/contrib/report/graph.py) refuses to build a plotly graph from an empty DataFrame. The constructor stores the df, and _init_data immediately raises ValueError('df is empty.') when df.empty is True, because there is nothing to plot and downstream slicing would silently produce blank figures.

Solutions

1. Inspect why the DataFrame is empty: check the upstream query (recorder.load_pred, analysis result, date range, instrument filter)
2. Widen or correct the date/instrument selection so at least one row matches
3. Guard at the call site: skip plotting when df.empty instead of constructing the graph
4. If predictions are missing, run signal recording (SignalRecord) before generating report graphs

Example fix

# before
graph = ScatterGraph(df=result_df)

# after
if result_df.empty:
    logger.warning('no data to plot; skip')
else:
    graph = ScatterGraph(df=result_df)

if df is None or df.empty:
    raise ValueError(f'cannot build graph: df has {0 if df is None else len(df)} rows')
graph = ScatterGraph(df=df)

def has_plot_data(df: pd.DataFrame) -> bool:
    return df is not None and not df.empty and len(df.columns) > 0

try:
    graph = ScatterGraph(df=df)
except ValueError as e:
    if 'empty' in str(e):
        logger.warning('skipping plot: input dataframe empty')
    else:
        raise

When it happens

Trigger: Passing an empty pd.DataFrame (zero rows or zero columns) as the df kwarg to any BaseGraph subclass (e.g. ScatterGraph, BarGraph) or to a graph built via BaseGraph.get_instance_with_graph_parameters with df=empty_df.

Common situations: Feeding the graph a recorder's prediction/analysis output that came back empty (e.g. no predictions were saved for the segment); filtering a report DataFrame by date/instrument conditions that match nothing; upstream data preparation silently returning an empty frame.

Related errors

• benchmark freq can't be None!
• Both trade_end_time and bench_value is None, benchmark is…
• Empty data from dataset, please check your dataset config.
• Empty data from dataset, please check your dataset config.
• Empty data from dataset, please check your dataset config.

AI-assisted analysis of microsoft/qlib@79633dd950 (2026-08-15). Data as JSON: /api/errors/8655825f3cf32812. Report an issue: GitHub.

Appendix: source

Thrown at qlib/contrib/report/graph.py:52

        """
        self._df = df

        self._layout = dict() if layout is None else layout
        self._graph_kwargs = dict() if graph_kwargs is None else graph_kwargs
        self._name_dict = name_dict

        self.data = None

        self._init_parameters(**kwargs)
        self._init_data()

    def _init_data(self):
        """

        :return:
        """
        if self._df.empty:
            raise ValueError("df is empty.")

        self.data = self._get_data()

    def _init_parameters(self, **kwargs):
        """

        :param kwargs
        """

        # Instantiate graphics parameters
        self._graph_type = self._name.lower().capitalize()

        # Displayed column name
        if self._name_dict is None:
            self._name_dict = {_item: _item for _item in self._df.columns}

    @staticmethod
    def get_instance_with_graph_parameters(graph_type: str = None, **kwargs):

View on GitHub (pinned to 79633dd950)