Skip to content

:hedgehog: PostHog is the leading platform for building self-driving products. Our developer tools – AI observability, analytics, session replay, flags, experiments, error tracking, logs, and more – capture all the context agents need to diagnose problems, uncover opportunities, and ship fixes. Steer it all from Slack, web, desktop, or the MCP.

1.6 KB
Raw Blame
from typing import Any

from .utils import format_matrix, format_number


class BoxPlotResultsFormatter:
    """
    Compresses and formats box plot results into a LLM-friendly string.

    Single series:
    ```
    Date|Min|P25|Median|P75|Max|Mean
    2025-01-20|1.2|5.5|12.3|25.8|100.4|18.7
    ```

    Multiple series:
    ```
    Date|Series|Min|P25|Median|P75|Max|Mean
    2025-01-20|$pageview|1.2|5.5|12.3|25.8|100.4|18.7
    ```
    """

    STATS: list[tuple[str, str]] = [
        ("Min", "min"),
        ("P25", "p25"),
        ("Median", "median"),
        ("P75", "p75"),
        ("Max", "max"),
        ("Mean", "mean"),
    ]

    def __init__(self, boxplot_data: list[dict[str, Any]]):
        self._data = boxplot_data

    def format(self) -> str:
        if not self._data:
            return "No data recorded for this time period."

        series_labels = {datum.get("series_label") for datum in self._data if datum.get("series_label") is not None}
        multi_series = len(series_labels) > 1

        stat_columns = [col for col, _key in self.STATS]

        matrix: list[list[str]] = []
        if multi_series:
            matrix.append(["Date", "Series", *stat_columns])
        else:
            matrix.append(["Date", *stat_columns])

        for datum in self._data:
            row = [datum["day"]]
            if multi_series:
                row.append(datum.get("series_label") or datum.get("label", ""))
            for _col, key in self.STATS:
                row.append(format_number(datum.get(key)))
            matrix.append(row)

        return format_matrix(matrix)