File size: 7,083 Bytes
b5186fd fccf1b6 b5186fd fccf1b6 b5186fd fccf1b6 b5186fd |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 |
from fastapi import FastAPI
from contextlib import asynccontextmanager
import joblib, os, requests, pandas as pd
from datetime import datetime
from typing import Literal, Annotated
from pydantic import BaseModel, Field
HF_REPO = "samithcs/heart-rate-models"
HEART_MODEL_FILENAME = "Heart_Rate_Predictor_model.joblib"
ANOMALY_MODEL_FILENAME = "Anomaly_Detector_model.joblib"
MODEL_DIR = os.path.join("artifacts", "model_trainer")
os.makedirs(MODEL_DIR, exist_ok=True)
def download_from_hf(filename):
local_path = os.path.join(MODEL_DIR, filename)
if os.path.exists(local_path):
print(f"✅ {filename} already exists at {local_path}")
return local_path
url = f"https://huggingface.co/{HF_REPO}/resolve/main/{filename}"
print(f"⬇️ Downloading {filename} from {url} ...")
with requests.get(url, stream=True) as r:
r.raise_for_status()
with open(local_path, "wb") as f:
for chunk in r.iter_content(chunk_size=8192):
f.write(chunk)
print(f"✅ Downloaded {filename} to {local_path}")
return local_path
# ===============================
# Lifespan context
# ===============================
@asynccontextmanager
async def lifespan(app: FastAPI):
global heart_model, heart_features, anomaly_model, anomaly_features
HEART_MODEL_PATH = download_from_hf(HEART_MODEL_FILENAME)
ANOMALY_MODEL_PATH = download_from_hf(ANOMALY_MODEL_FILENAME)
heart_model_artifacts = joblib.load(HEART_MODEL_PATH)
heart_model = heart_model_artifacts['model']
heart_features = heart_model_artifacts['feature_columns']
anomaly_model_artifacts = joblib.load(ANOMALY_MODEL_PATH)
anomaly_model = anomaly_model_artifacts['model']
anomaly_features = anomaly_model_artifacts['feature_columns']
yield
# ===============================
# FastAPI app
# ===============================
app = FastAPI(title="Health Monitoring API", lifespan=lifespan)
# ===============================
# Request schemas
# ===============================
class HeartRateInput(BaseModel):
age: Annotated[int, Field(..., gt=0, lt=120)]
gender: Annotated[Literal['M', 'F'], Field(...)]
weight_kg: Annotated[float, Field(..., gt=0)]
height_cm: Annotated[float, Field(..., gt=0, lt=250)]
bmi: Annotated[float, Field(..., gt=0, lt=100)]
fitness_level: Annotated[Literal['lightly_active','fairly_active','sedentary','very_active'], Field(...)]
performance_level: Annotated[Literal['low','moderate','high'], Field(...)]
resting_hr: Annotated[int, Field(..., gt=0, lt=120)]
max_hr: Annotated[int, Field(..., gt=0, lt=220)]
activity_type: Annotated[Literal['sleeping','walking','resting','light','commuting','exercise'], Field(...)]
activity_intensity: Annotated[float, Field(..., gt=0.0)]
steps_5min: Annotated[int, Field(..., gt=0)]
calories_5min: Annotated[float, Field(..., gt=0)]
hrv_rmssd: Annotated[float, Field(..., gt=0)]
stress_score: Annotated[int, Field(..., gt=0, lt=100)]
signal_quality: Annotated[float, Field(..., gt=0)]
skin_temperature: Annotated[float, Field(..., gt=0)]
device_battery: Annotated[int, Field(..., gt=0)]
elevation_gain: Annotated[int, Field(..., ge=0)]
sleep_stage: Annotated[Literal['light_sleep','deep_sleep','rem_sleep'], Field(...)]
date: Annotated[datetime, Field(...)]
class AnomalyInput(BaseModel):
heart_rate: Annotated[float, Field(..., gt=0.0)]
resting_hr_baseline: Annotated[int, Field(..., gt=0, lt=120)]
activity_type: Annotated[Literal['sleeping','walking','resting','light','commuting','exercise'], Field(...)]
activity_intensity: Annotated[float, Field(..., gt=0)]
steps_5min: Annotated[int, Field(..., gt=0)]
calories_5min: Annotated[float, Field(..., gt=0)]
hrv_rmssd: Annotated[float, Field(..., gt=0)]
stress_score: Annotated[int, Field(..., gt=0, lt=100)]
confidence_score: Annotated[float, Field(..., gt=0.0)]
signal_quality: Annotated[float, Field(..., gt=0)]
skin_temperature: Annotated[float, Field(..., gt=0)]
device_battery: Annotated[int, Field(..., gt=0)]
elevation_gain: Annotated[int, Field(..., ge=0)]
sleep_stage: Annotated[Literal['light_sleep','deep_sleep','rem_sleep'], Field(...)]
date: Annotated[datetime, Field(...)]
# ===============================
# Startup event to download & load models
# ===============================
@app.on_event("startup")
def startup_event():
global heart_model, heart_features, anomaly_model, anomaly_features
HEART_MODEL_PATH = download_from_hf(HEART_MODEL_FILENAME)
ANOMALY_MODEL_PATH = download_from_hf(ANOMALY_MODEL_FILENAME)
heart_model_artifacts = joblib.load(HEART_MODEL_PATH)
heart_model = heart_model_artifacts['model']
heart_features = heart_model_artifacts['feature_columns']
anomaly_model_artifacts = joblib.load(ANOMALY_MODEL_PATH)
anomaly_model = anomaly_model_artifacts['model']
anomaly_features = anomaly_model_artifacts['feature_columns']
# ===============================
# Utility: preprocess features
# ===============================
def preprocess_heart_features(data_dict: dict) -> pd.DataFrame:
data_dict['date_encoded'] = data_dict['date'].timestamp()
data_dict['gender_M'] = 1 if data_dict['gender']=='M' else 0
data_dict['gender_F'] = 1 if data_dict['gender']=='F' else 0
for act in ['sleeping','walking','resting','light','commuting','exercise']:
data_dict[f"activity_type_{act}"] = 1 if data_dict['activity_type']==act else 0
for stage in ['light_sleep','deep_sleep','rem_sleep']:
data_dict[f"sleep_stage_{stage}"] = 1 if data_dict['sleep_stage']==stage else 0
return pd.DataFrame([{f: data_dict.get(f,0) for f in heart_features}])
def preprocess_anomaly_features(data_dict: dict) -> pd.DataFrame:
data_dict['date_encoded'] = data_dict['date'].timestamp()
for act in ['sleeping','walking','resting','light','commuting','exercise']:
data_dict[f"activity_type_{act}"] = 1 if data_dict['activity_type']==act else 0
for stage in ['light_sleep','deep_sleep','rem_sleep']:
data_dict[f"sleep_stage_{stage}"] = 1 if data_dict['sleep_stage']==stage else 0
return pd.DataFrame([{f: data_dict.get(f,0) for f in anomaly_features}])
# ===============================
# Endpoints
# ===============================
@app.get("/")
def home():
return {"message":"Health Monitoring API is running!"}
@app.post("/predict_heart_rate")
def predict_heart_rate(input_data: HeartRateInput):
try:
X = preprocess_heart_features(input_data.model_dump())
prediction = heart_model.predict(X)[0]
return {"heart_rate_prediction": float(prediction)}
except Exception as e:
return {"error": str(e)}
@app.post("/detect_anomaly")
def detect_anomaly(input_data: AnomalyInput):
try:
X = preprocess_anomaly_features(input_data.model_dump())
prediction = anomaly_model.predict(X)[0]
return {"anomaly_detected": bool(prediction)}
except Exception as e:
return {"error": str(e)}
|