{% if summary["associations_skipped"] %}

Computing pairwise associations was skipped. This is due to either:

The dataframe exceeding the configured table_report_associations_threshold limit (default: 30).
The compute_associations option being set to False (default: "auto", which applies the configured table_report_associations_threshold).

You can adjust this behavior in several ways:

To force computation for a single report:

    report = TableReport(df, compute_associations=True)

To change the threshold for the current Python session, use skrub.set_config:

    from skrub import set_config
    set_config(table_report_associations_threshold=50)

To make the change permanent, use an environment variable:
```
    export SKB_TABLE_REPORT_ASSOCIATIONS_THRESHOLD=50
```

{% elif summary.get("associations_skipped_polars_no_pyarrow", False) %}

Computing pairwise associations is not available for Polars dataframes when PyArrow is not installed. To enable associations, install PyArrow: pip install pyarrow, or use a Pandas dataframe.

{% elif summary["top_associations"] %}

{% for association in summary["top_associations"] %} {% endfor %}

Column 1	Column 2	Cramér's V	Pearson's Correlation
{{ association["left_column_name"] }}	{{ association["right_column_name"] }}	{{ association["cramer_v"] \| format_number }}	{%- if not (association["pearson_corr"] \| is_null) -%} {{ association["pearson_corr"] \| format_number }} {%- endif -%}

The table below shows the strength of association between the most similar columns in the dataframe. Cramér's V statistic is a number between 0 and 1. When it is close to 1 the columns are strongly associated — they contain similar information. In this case, one of them may be redundant and for some models (such as linear models) it might be beneficial to remove it.

{% else %} No strong associations between any pair of columns were identified by a quick screening of a subsample of the dataframe. {% endif %}