DEV-61 Recommender system - minor imporvments

models/recommendation.py

@@ -1,37 +1,182 @@
+"""
+Plant recommendation engine with fuzzy matching and optional
+light/temperature data import from .bin / .npz files.
+"""
+
+import numpy as np
+import os
 from database.plants import df_plants
 
+# ----------------------------------------------------------------------
+#  Original fuzzy‑matching helpers
+# ----------------------------------------------------------------------
 def _triangular_membership(x, center, spread):
-    if spread <= 0: return 1.0 if x == center else 0.0
-    return max(0.0, 1.0 - abs(x-center)/spread)
+    if spread <= 0:
+        return 1.0 if x == center else 0.0
+    return max(0.0, 1.0 - abs(x - center) / spread)
 
 def _fuzzy_match(user_val, plant_val, spread):
     return _triangular_membership(plant_val, user_val, spread)
 
+# ----------------------------------------------------------------------
+#  Public recommendation function
+# ----------------------------------------------------------------------
 def recommend_plants(user_prefs, top_n=5):
-    if df_plants is None: return []
+    """
+    Return a list of (plant_name, score) tuples sorted by score.
+    user_prefs is a dict that may contain:
+        water, sunlight, temp, pet_safe, space,
+        allergy_concern, existing_plants
+    """
+    if df_plants is None:
+        return []
+
     scores = []
     for _, plant in df_plants.iterrows():
-        w_match = _fuzzy_match(user_prefs["water"],    plant["water"],       2.0)
-        s_match = _fuzzy_match(user_prefs["sunlight"], plant["sunlight"],    2.0)
-        t_match = _fuzzy_match(user_prefs["temp"],     plant["temperature"], 4.0)
-        p_match = (1.0 if plant["pet_safe"] == user_prefs["pet_safe"] else 0.0) if user_prefs.get("pet_safe") is not None else None
-        space_match = (1.0 if plant["space"] == user_prefs["space"] else 0.0) if user_prefs.get("space") is not None else None
+        w_match = _fuzzy_match(user_prefs["water"], plant["water"], 2.0)
+        s_match = _fuzzy_match(user_prefs["sunlight"], plant["sunlight"], 2.0)
+        t_match = _fuzzy_match(user_prefs["temp"], plant["temperature"], 4.0)
+
+        p_match = (1.0 if plant["pet_safe"] == user_prefs["pet_safe"] else 0.0) \
+                  if user_prefs.get("pet_safe") is not None else None
+
+        space_match = (1.0 if plant["space"] == user_prefs["space"] else 0.0) \
+                      if user_prefs.get("space") is not None else None
+
         if user_prefs.get("allergy_concern") is not None:
-            a_match = (1.0 if not plant["pollen_allergies"] else 0.0) if user_prefs["allergy_concern"] else 1.0
+            a_match = (1.0 if not plant["pollen_allergies"] else 0.0) \
+                      if user_prefs["allergy_concern"] else 1.0
         else:
             a_match = None
+
         user_existing = user_prefs.get("existing_plants", [])
         if user_existing:
             plant_compat_list = plant["existing_plants"]
-            exist_match = (sum(1 for up in user_existing if up in plant_compat_list)/len(user_existing)) if plant_compat_list else 0.0
+            exist_match = (sum(1 for up in user_existing if up in plant_compat_list) /
+                           len(user_existing)) if plant_compat_list else 0.0
         else:
             exist_match = None
-        weights = {"water":0.15,"sunlight":0.15,"temp":0.15,"pet":0.15,"space":0.15,"allergy":0.15,"existing":0.10}
-        components = {"water":w_match,"sunlight":s_match,"temp":t_match,"pet":p_match,"space":space_match,"allergy":a_match,"existing":exist_match}
-        active = {k:v for k,v in components.items() if v is not None}
-        if not active: continue
+
+        weights = {
+            "water": 0.15, "sunlight": 0.15, "temp": 0.15,
+            "pet": 0.15, "space": 0.15, "allergy": 0.15, "existing": 0.10
+        }
+        components = {
+            "water": w_match, "sunlight": s_match, "temp": t_match,
+            "pet": p_match, "space": space_match, "allergy": a_match,
+            "existing": exist_match
+        }
+        active = {k: v for k, v in components.items() if v is not None}
+        if not active:
+            continue
+
         active_weight_sum = sum(weights[k] for k in active)
-        score = sum(v*weights[k]/active_weight_sum for k,v in active.items())
+        score = sum(v * weights[k] / active_weight_sum for k, v in active.items())
         scores.append((plant["name"], score))
+
     scores.sort(key=lambda x: x[1], reverse=True)
-    return scores[:top_n]
+    return scores[:top_n]
+
+# ----------------------------------------------------------------------
+#  New helper: map lux readings to a sunlight score (1–10)
+# ----------------------------------------------------------------------
+def _lux_to_sunlight(lux):
+    """
+    Convert an average lux value into a plant‑care sunlight scale 1–10.
+    Thresholds are loosely based on indoor light levels.
+    """
+    # Piecewise linear mapping:
+    thresholds = [
+        (0,     100,   1.0,  2.0),
+        (100,   250,   2.0,  3.0),
+        (250,   500,   3.0,  4.0),
+        (500,   1000,  4.0,  5.0),
+        (1000,  2000,  5.0,  6.0),
+        (2000,  4000,  6.0,  7.0),
+        (4000,  8000,  7.0,  8.0),
+        (8000,  16000, 8.0,  9.0),
+        (16000, 32000, 9.0,  10.0),
+        (32000, float("inf"), 10.0, 10.0),
+    ]
+    for low, high, low_s, high_s in thresholds:
+        if low <= lux <= high:
+            if high == float("inf"):
+                return low_s
+            fraction = (lux - low) / (high - low)
+            return low_s + fraction * (high_s - low_s)
+    return 5.0   # fallback
+
+# ----------------------------------------------------------------------
+#  File‑analysis entry point
+# ----------------------------------------------------------------------
+def analyze_environment_file(file_path):
+    """
+    Read a .bin or .npz file that may contain light and/or temperature data.
+
+    Returns a dict with:
+        - "sunlight" : float (1‑10) derived from average light
+        - "temp"     : float or None (average temperature in °C)
+    """
+    if not os.path.exists(file_path):
+        raise FileNotFoundError(f"File not found: {file_path}")
+
+    ext = os.path.splitext(file_path)[1].lower()
+    light_values = []
+    temp_values  = []
+
+    try:
+        if ext == ".npz":
+            with np.load(file_path, allow_pickle=True) as npz:
+                # Look for known keys
+                for key in npz.files:
+                    arr = npz[key].flatten()
+                    if key.lower() in ("light", "lux", "illuminance"):
+                        light_values.extend(arr.tolist())
+                    elif key.lower() in ("temp", "temperature", "celsius"):
+                        temp_values.extend(arr.tolist())
+                # If no known keys, treat everything as light (back‑compat with dashboard)
+                if not light_values and not temp_values:
+                    for key in npz.files:
+                        light_values.extend(npz[key].flatten().tolist())
+
+        elif ext == ".bin":
+            raw = np.fromfile(file_path, dtype=np.float32)
+            light_values = raw.tolist()
+        else:
+            raise ValueError("Unsupported file type. Use .bin or .npz")
+
+    except Exception as e:
+        raise RuntimeError(f"Failed to read environment file: {e}")
+
+    # Compute sunlight from light data
+    sunlight = None
+    if light_values:
+        avg_lux = sum(light_values) / len(light_values)
+        sunlight = _lux_to_sunlight(avg_lux)
+    else:
+        sunlight = 5.0   # neutral guess if no light data
+
+    # Compute temperature if available
+    temp = None
+    if temp_values:
+        avg_temp = sum(temp_values) / len(temp_values)
+        temp = round(avg_temp, 1)
+
+    return {"sunlight": sunlight, "temp": temp}
+
+# ----------------------------------------------------------------------
+#  Convenience function that merges file‑derived values into prefs
+# ----------------------------------------------------------------------
+def merge_file_prefs(user_prefs, file_path):
+    """
+    Return a new user_prefs dict where 'sunlight' and 'temp' are
+    replaced with values from the file, if available.
+    Other keys are left untouched.
+    """
+    env = analyze_environment_file(file_path)
+    merged = dict(user_prefs)  # shallow copy
+    merged["sunlight"] = env["sunlight"]
+    if env["temp"] is not None:
+        merged["temp"] = env["temp"]
+    # If no temperature was found, keep the original temp value
+    return merged

ui/pages/recommendation.py

@@ -1,11 +1,13 @@
 import tkinter as tk
+from tkinter import filedialog, messagebox
+import os
 
 from .theme import (
     BG_MAIN, BG_CARD, BG_CARD2, BG_GLASS, ACCENT, ACCENT2, BLUE, RED, YELLOW, TEXT_PRI,
     TEXT_SEC, TEXT_MUT, BORDER, ON_ACCENT, bind_tree, hover, GreenSlider, BasePage
 )
 from database.plants import df_plants as _df_plants, PANDAS_OK
-from models.recommendation import recommend_plants
+from models.recommendation import recommend_plants, analyze_environment_file
 
 
 class RecommendationSystemPage(BasePage):
@@ -59,6 +61,24 @@ def slider_row(parent, label, from_, to, resolution, default):
         self._sunlight_var = slider_row(left_inner, "Sunlight (1–10)",    1, 10, 0.5, 6)
         self._temp_var     = slider_row(left_inner, "Temperature (°C)",  10, 40, 0.5, 22)
 
+        # ----- File import for light / temperature (overrides sliders) -----
+        file_frame = tk.Frame(left_inner, bg=BG_CARD)
+        file_frame.pack(fill="x", pady=(4, 0))
+        tk.Label(file_frame, text="Or load from light data file:",
+                 font=self.f_small, bg=BG_CARD, fg=TEXT_SEC).pack(anchor="w")
+        btn_row = tk.Frame(file_frame, bg=BG_CARD); btn_row.pack(fill="x", pady=(4,0))
+        file_btn = tk.Frame(btn_row, bg=ACCENT, cursor="hand2", padx=12, pady=5)
+        file_btn.pack(side="left")
+        tk.Label(file_btn, text="📂 Choose File", font=self.f_label,
+                 bg=ACCENT, fg=ON_ACCENT).pack()
+        bind_tree(file_btn, "<Button-1>", lambda e: self._load_light_file())
+        hover(file_btn, ACCENT, ACCENT2)
+        self._file_info_var = tk.StringVar(value="")
+        tk.Label(btn_row, textvariable=self._file_info_var,
+                 font=("Segoe UI",8), bg=BG_CARD, fg=TEXT_MUT,
+                 wraplength=200, justify="left").pack(side="left", padx=(8,0))
+        # ------------------------------------------------------------------
+
         tk.Frame(left_inner, bg=BORDER, height=1).pack(fill="x", pady=(8,10))
 
         # Space selector
@@ -138,6 +158,28 @@ def toggle(e, v=var, l=lbl):
         self._results_frame.pack(fill="both", expand=True)
         self._show_finder_empty()
 
+    def _load_light_file(self):
+        """Let the user pick a .bin or .npz file and update sunlight / temp sliders."""
+        path = filedialog.askopenfilename(
+            title="Select Light Data File",
+            filetypes=[("Binary/NumPy files", "*.bin *.npz"), ("All files", "*.*")]
+        )
+        if not path:
+            return
+        try:
+            env = analyze_environment_file(path)
+            # Update sliders – values will be clamped to their ranges automatically
+            self._sunlight_var.set(env["sunlight"])
+            if env["temp"] is not None:
+                self._temp_var.set(env["temp"])
+            # Show what was loaded
+            temp_str = f"{env['temp']}°C" if env["temp"] is not None else "—"
+            self._file_info_var.set(
+                f"✔ {os.path.basename(path)}\nLight: {env['sunlight']:.1f}  Temp: {temp_str}"
+            )
+        except Exception as e:
+            messagebox.showerror("File Error", f"Could not read file:\n{e}")
+
     def _show_finder_empty(self):
         for w in self._results_frame.winfo_children(): w.destroy()
         container = tk.Frame(self._results_frame, bg=BG_CARD, padx=20, pady=40)