add Trends page: GMI trend, overnight/daytime, weekday pattern, bolus, site-age (FR7, FR8)

reporting.py gains overnight_daytime_split, day_of_week_pattern,
bolus_stats, and site_interval_stats. The last two are explicitly scoped
to bolus-only insulin (no basal reconstruction) since that needs the
Nightscout profile, which is the separate profile-tuning feature's
territory — documented as a caveat in both the UI and the design doc.

Co-Authored-By: Claude Sonnet 5 <[email protected]>

reporting.py

@@ -9,8 +9,12 @@ definities waarop dit gebaseerd is (Battelino et al. 2019, Bergenstal et al.
 
 from __future__ import annotations
 
+from itertools import pairwise
+
 import pandas as pd
 
+from nightscout_client import INSULIN_EVENT_TYPES
+
 # Klinische drempels in mmol/L (Battelino et al. 2019, Diabetes Care 42(8):1593-1603).
 VERY_LOW = 3.0
 LOW = 3.9
@@ -226,3 +230,132 @@ def agp_percentile_bands(glucose: pd.DataFrame, bin_minutes: int = 30) -> pd.Dat
     grouped.columns = ["p5", "p25", "p50", "p75", "p95"]
     grouped = grouped.reset_index().rename(columns={"bin": "minute_of_day"})
     return grouped[AGP_PERCENTILE_COLUMNS].sort_values("minute_of_day").reset_index(drop=True)
+
+
+def overnight_daytime_split(
+    glucose: pd.DataFrame, overnight_start: int = 0, overnight_end: int = 6
+) -> dict[str, dict]:
+    """Tijd-in-bereik apart voor de nacht (overnight_start-overnight_end uur) en overdag."""
+    empty = tir_breakdown_5band(glucose)
+    if glucose.empty:
+        return {"overnight": empty, "daytime": empty}
+
+    hour = glucose["timestamp"].dt.hour
+    is_overnight = (hour >= overnight_start) & (hour < overnight_end)
+    return {
+        "overnight": tir_breakdown_5band(glucose[is_overnight]),
+        "daytime": tir_breakdown_5band(glucose[~is_overnight]),
+    }
+
+
+WEEKDAY_NAMES = ["Maandag", "Dinsdag", "Woensdag", "Donderdag", "Vrijdag", "Zaterdag", "Zondag"]
+
+
+def day_of_week_pattern(glucose: pd.DataFrame) -> pd.DataFrame:
+    """Gemiddelde glucose en tijd-in-bereik per weekdag."""
+    columns = ["weekday", "mean_mmol", "in_range_pct", "count"]
+    if glucose.empty:
+        return pd.DataFrame(columns=columns)
+
+    df = glucose.copy()
+    df["weekday"] = df["timestamp"].dt.dayofweek
+
+    rows = []
+    for weekday in range(7):
+        day_df = df[df["weekday"] == weekday]
+        if day_df.empty:
+            rows.append(
+                {"weekday": WEEKDAY_NAMES[weekday], "mean_mmol": None, "in_range_pct": None, "count": 0}
+            )
+            continue
+        tir = tir_breakdown_5band(day_df)
+        rows.append(
+            {
+                "weekday": WEEKDAY_NAMES[weekday],
+                "mean_mmol": round(day_df["mmol"].mean(), 1),
+                "in_range_pct": tir["in_range"],
+                "count": len(day_df),
+            }
+        )
+    return pd.DataFrame(rows, columns=columns)
+
+
+def bolus_stats(treatments: pd.DataFrame) -> dict:
+    """Samenvatting van bolus-insuline, opgesplitst naar type.
+
+    Let op: dit betreft alleen bolusinsuline (maaltijd/correctie/combo/SMB).
+    Basaalinsuline wordt hier niet in meegenomen — dat vereist profielgegevens
+    (zie docs/design-reporting.md, scoping-kanttekening) en levert dus geen
+    volledige totale-dagdosis op.
+    """
+    if treatments.empty:
+        return {"total_units": 0.0, "count": 0, "by_type": {}}
+
+    bolus = treatments[treatments["eventType"].isin(INSULIN_EVENT_TYPES) & treatments["insulin"].notna()]
+    if bolus.empty:
+        return {"total_units": 0.0, "count": 0, "by_type": {}}
+
+    by_type = bolus.groupby("eventType")["insulin"].sum().round(1).to_dict()
+    return {
+        "total_units": round(bolus["insulin"].sum(), 1),
+        "count": len(bolus),
+        "by_type": by_type,
+    }
+
+
+SITE_INTERVAL_COLUMNS = [
+    "start",
+    "end",
+    "duration_days",
+    "mean_mmol",
+    "in_range_pct",
+    "total_bolus_units",
+    "avg_bolus_units_per_day",
+    "total_carbs_g",
+]
+
+
+def _site_interval_row(glucose: pd.DataFrame, treatments: pd.DataFrame, start, end) -> dict:
+    duration_days = max((end - start).total_seconds() / 86400, 1e-9)
+    window_glucose = glucose[(glucose["timestamp"] >= start) & (glucose["timestamp"] < end)]
+    window_treatments = (
+        treatments[(treatments["timestamp"] >= start) & (treatments["timestamp"] < end)]
+        if not treatments.empty
+        else treatments
+    )
+    bolus = bolus_stats(window_treatments)
+    total_carbs = window_treatments["carbs"].fillna(0).sum() if not treatments.empty else 0.0
+    tir = tir_breakdown_5band(window_glucose)
+
+    return {
+        "start": start,
+        "end": end,
+        "duration_days": round(duration_days, 1),
+        "mean_mmol": round(window_glucose["mmol"].mean(), 1) if not window_glucose.empty else None,
+        "in_range_pct": tir["in_range"],
+        "total_bolus_units": bolus["total_units"],
+        "avg_bolus_units_per_day": round(bolus["total_units"] / duration_days, 1),
+        "total_carbs_g": round(total_carbs, 0),
+    }
+
+
+def site_interval_stats(
+    glucose: pd.DataFrame, treatments: pd.DataFrame, site_changes: pd.DataFrame
+) -> pd.DataFrame:
+    """Statistieken per interval tussen opeenvolgende pompwissels (FR8).
+
+    Voor elk interval sinds een pompwissel: duur in dagen ("leeftijd" van de
+    pompsite bij het einde van het interval), gemiddelde glucose,
+    tijd-in-bereik, en bolus-insuline/koolhydraten in dat interval — als
+    benadering van of insulinebehoefte samenhangt met de leeftijd van de
+    pompsite. Zie bolus_stats() voor de kanttekening dat basaalinsuline hier
+    niet in zit.
+    """
+    if site_changes.empty or glucose.empty:
+        return pd.DataFrame(columns=SITE_INTERVAL_COLUMNS)
+
+    change_times = sorted(site_changes["timestamp"])
+    boundaries = [*change_times, glucose["timestamp"].max()]
+
+    rows = [_site_interval_row(glucose, treatments, start, end) for start, end in pairwise(boundaries)]
+    return pd.DataFrame(rows, columns=SITE_INTERVAL_COLUMNS)