bolade
808 lines
30 KiB
Plaintext
808 lines
30 KiB
Plaintext
{
|
||
"cells": [
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 2,
|
||
"id": "b18c1027",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"{'id': 'gen-1759135172-DIhs7TMuaaVY0h3T2ibV', 'provider': 'Google', 'model': 'google/gemini-2.5-flash-lite', 'object': 'chat.completion', 'created': 1759135172, 'choices': [{'logprobs': None, 'finish_reason': 'stop', 'native_finish_reason': 'STOP', 'index': 0, 'message': {'role': 'assistant', 'content': 'Parameters,Best,LLN,Pred.,%Pred.,ZScore,PRE#1,PRE#2,PRE#3\\nFVC,L,4.24,3.03,3.79,112.0,0.95,4.24,4.17,4.15\\nFEV1,L,3.26,2.53,3.16,103.3,0.28,3.26,3.21,3.14\\nFEV1/FVC%,76.89,72.47,83.78,91.8,-1.05,76.9,77.0,75.7\\nPEF,L/m,684,222,384,178.7,-,444,438,684\\nFEF2575,L/s,2.74,2.15,3.42,80.2,-0.84,2.74,2.68,2.48\\nFEF25,L/s,6.08,-,-,-,6.08,6.0,5.53\\nFEF50,L/s,3.06,-,-,-,3.06,3.1,2.77\\nFEF75,L/s,1.06,0.71,1.41,75.1,-0.72,1.06,1.12,0.94\\nPEFTime,ms,-,-,79,-,79,49,39\\nEvol,mL,-,-,78.0,-,78.0,77.0,197.0\\nFEV6,L,4.22,3.03,3.79,111.4,-,4.22,4.17,4.13', 'refusal': None, 'reasoning': None}}], 'usage': {'prompt_tokens': 1350, 'completion_tokens': 454, 'total_tokens': 1804, 'prompt_tokens_details': {'cached_tokens': 0}, 'completion_tokens_details': {'reasoning_tokens': 0, 'image_tokens': 0}}}\n",
|
||
"Content saved to extracted_table.csv\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"\n",
|
||
"import requests\n",
|
||
"import json\n",
|
||
"import base64\n",
|
||
"from pathlib import Path\n",
|
||
"\n",
|
||
"API_KEY_REF = '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'\n",
|
||
"def encode_pdf_to_base64(pdf_path):\n",
|
||
" with open(pdf_path, \"rb\") as pdf_file:\n",
|
||
" return base64.b64encode(pdf_file.read()).decode('utf-8')\n",
|
||
"\n",
|
||
"url = \"https://openrouter.ai/api/v1/chat/completions\"\n",
|
||
"headers = {\n",
|
||
" \"Authorization\": f\"Bearer {API_KEY_REF}\",\n",
|
||
" \"Content-Type\": \"application/json\"\n",
|
||
"}\n",
|
||
"\n",
|
||
"# Read and encode the PDF\n",
|
||
"pdf_path = \"data/~Moran~K~19910201~Spirometry Exam~20250729~20250729032843.pdf\"\n",
|
||
"base64_pdf = encode_pdf_to_base64(pdf_path)\n",
|
||
"data_url = f\"data:application/pdf;base64,{base64_pdf}\"\n",
|
||
"\n",
|
||
"messages = [\n",
|
||
" {\n",
|
||
" \"role\": \"user\",\n",
|
||
" \"content\": [\n",
|
||
" {\n",
|
||
" \"type\": \"text\",\n",
|
||
" \"text\": \"Please extract the Spirometry table from the pdf and return the values in csv format, \"\n",
|
||
" \"note that it is the unit of parameter that is beside it and it should not be a column. \"\n",
|
||
" \"The '-' Should be treated as empty values.\"\n",
|
||
" \"do not add 'csv' at the start or end of the response\"\n",
|
||
" },\n",
|
||
" {\n",
|
||
" \"type\": \"file\",\n",
|
||
" \"file\": {\n",
|
||
" \"filename\": \"document.pdf\",\n",
|
||
" \"file_data\": data_url\n",
|
||
" }\n",
|
||
" },\n",
|
||
" ]\n",
|
||
" }\n",
|
||
"]\n",
|
||
"\n",
|
||
"# Optional: Configure PDF processing engine\n",
|
||
"# PDF parsing will still work even if the plugin is not explicitly set\n",
|
||
"plugins = [\n",
|
||
" {\n",
|
||
" \"id\": \"file-parser\",\n",
|
||
" \"pdf\": {\n",
|
||
" \"engine\": \"pdf-text\" # defaults to \"mistral-ocr\". See Pricing above\n",
|
||
" }\n",
|
||
" }\n",
|
||
"]\n",
|
||
"\n",
|
||
"payload = {\n",
|
||
" \"model\": \"google/gemini-2.5-flash-lite\",\n",
|
||
" \"messages\": messages,\n",
|
||
"}\n",
|
||
"\n",
|
||
"response = requests.post(url, headers=headers, json=payload)\n",
|
||
"# Get the response content\n",
|
||
"response_data = response.json()\n",
|
||
"print(response_data)\n",
|
||
"\n",
|
||
"# Extract the content from the response\n",
|
||
"if 'choices' in response_data and len(response_data['choices']) > 0:\n",
|
||
" content = response_data['choices'][0]['message']['content']\n",
|
||
" \n",
|
||
" # Save to a CSV file\n",
|
||
" output_file = \"extracted_table.csv\"\n",
|
||
" with open(output_file, 'w', encoding='utf-8') as f:\n",
|
||
" f.write(content)\n",
|
||
" \n",
|
||
" print(f\"Content saved to {output_file}\")\n",
|
||
"else:\n",
|
||
" print(\"No content found in response\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 12,
|
||
"id": "56a9d655",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"FVC Best: 4.24, FVC Pred: 112.0\n",
|
||
"FEV1 Best: 3.26, FEV1 Pred: 103.3\n",
|
||
"FEV1/FVC% Best: 76.89, FEV1/FVC% Pred: 91.8\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"import pandas as pd\n",
|
||
"spirometry_df = pd.read_csv(\"extracted_table.csv\")\n",
|
||
"\n",
|
||
"fvc_best = spirometry_df.loc[spirometry_df['Parameters'] == 'FVC', 'Best'].values[0]\n",
|
||
"fvc_pred = spirometry_df.loc[spirometry_df['Parameters'] == 'FVC', '%Pred.'].values[0]\n",
|
||
"\n",
|
||
"fev1_best = spirometry_df.loc[spirometry_df['Parameters'] == 'FEV1', 'Best'].values[0]\n",
|
||
"fev1_pred = spirometry_df.loc[spirometry_df['Parameters'] == 'FEV1', '%Pred.'].values[0]\n",
|
||
"\n",
|
||
"fev1_fevc_best = spirometry_df.loc[spirometry_df['Parameters'] == 'FEV1/FVC%', 'Best'].values[0]\n",
|
||
"fev1_fevc_pred = spirometry_df.loc[spirometry_df['Parameters'] == 'FEV1/FVC%', '%Pred.'].values[0]\n",
|
||
"\n",
|
||
"print(f\"FVC Best: {fvc_best}, FVC Pred: {fvc_pred}\")\n",
|
||
"print(f\"FEV1 Best: {fev1_best}, FEV1 Pred: {fev1_pred}\")\n",
|
||
"print(f\"FEV1/FVC% Best: {fev1_fevc_best}, FEV1/FVC% Pred: {fev1_fevc_pred}\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 16,
|
||
"id": "990f4b4f",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Peak VT: 2.75\n",
|
||
"HR at Peak VT: 155.0\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"df = pd.read_csv('data/Pnoe_20250729_1550-Moran_Keirstyn.csv', delimiter=';')\n",
|
||
"peak_vt = df['VT(l)'].max()\n",
|
||
"max_vt_row = df.loc[df['VT(l)'].idxmax()]\n",
|
||
"print(f\"Peak VT: {peak_vt}\")\n",
|
||
"hr = max_vt_row['HR(bpm)']\n",
|
||
"print(f\"HR at Peak VT: {hr}\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 18,
|
||
"id": "041cbc3d",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Peak VT: 2.3770000000000002\n",
|
||
"HR at Peak VT: 171.525\n"
|
||
]
|
||
},
|
||
{
|
||
"name": "stderr",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"/tmp/ipykernel_301535/4157056299.py:3: FutureWarning: errors='ignore' is deprecated and will raise in a future version. Use to_numeric without passing `errors` and catch exceptions explicitly instead\n",
|
||
" df = df.apply(pd.to_numeric, errors='ignore')\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"df = pd.read_csv('data/Pnoe_20250729_1550-Moran_Keirstyn.csv', delimiter=';')\n",
|
||
"# Convert all columns to numeric where possible, coercing errors to NaN\n",
|
||
"df = df.apply(pd.to_numeric, errors='ignore')\n",
|
||
"df['VO2 Pulse'] = df['VO2(ml/min)'] / df['HR(bpm)'] # VO2 Pulse in mL/beat\n",
|
||
"df['VO2 Breath'] = df['VO2(ml/min)'] / df['BF(bpm)'] # VO2 per Breath in mL/breath\n",
|
||
"df['CHO'] = df['EE(kcal/min)'] * df['CARBS(%)']/100\n",
|
||
"df['FAT'] = df['EE(kcal/min)'] * df['FAT(%)']/100\n",
|
||
"# Smooth key columns using rolling window\n",
|
||
"window_size = 10\n",
|
||
"\n",
|
||
"# List of columns to smooth\n",
|
||
"columns_to_smooth = ['VO2(ml/min)', 'VCO2(ml/min)', 'HR(bpm)', 'VT(l)', 'BF(bpm)', 'VE(l/min)', 'VO2 Pulse', 'VO2 Breath', 'CHO', 'FAT']\n",
|
||
"\n",
|
||
"# Apply smoothing to each column\n",
|
||
"for col in columns_to_smooth:\n",
|
||
" if col in df.columns:\n",
|
||
" df[f'{col}_smoothed'] = df[col].rolling(window=window_size).mean()\n",
|
||
" \n",
|
||
"peak_vt = df['VT(l)_smoothed'].max()\n",
|
||
"max_vt_row = df.loc[df['VT(l)_smoothed'].idxmax()]\n",
|
||
"print(f\"Peak VT: {peak_vt}\")\n",
|
||
"hr = max_vt_row['HR(bpm)_smoothed']\n",
|
||
"print(f\"HR at Peak VT: {hr}\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 20,
|
||
"id": "de7cadd1",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Percent FEV: 72.91411042944786\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"percent_fev = (peak_vt / fev1_best) * 100\n",
|
||
"print(f\"Percent FEV: {percent_fev}\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 24,
|
||
"id": "cb972ed3",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/html": [
|
||
"<div>\n",
|
||
"<style scoped>\n",
|
||
" .dataframe tbody tr th:only-of-type {\n",
|
||
" vertical-align: middle;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe tbody tr th {\n",
|
||
" vertical-align: top;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe thead th {\n",
|
||
" text-align: right;\n",
|
||
" }\n",
|
||
"</style>\n",
|
||
"<table border=\"1\" class=\"dataframe\">\n",
|
||
" <thead>\n",
|
||
" <tr style=\"text-align: right;\">\n",
|
||
" <th></th>\n",
|
||
" <th>MeasurementDate</th>\n",
|
||
" <th>Comment</th>\n",
|
||
" <th>ExternalDeviceId</th>\n",
|
||
" <th>ExternalPatientId</th>\n",
|
||
" <th>FirstName</th>\n",
|
||
" <th>LastName</th>\n",
|
||
" <th>BirthDate</th>\n",
|
||
" <th>Age</th>\n",
|
||
" <th>Ethnicity</th>\n",
|
||
" <th>Gender</th>\n",
|
||
" <th>...</th>\n",
|
||
" <th>Child_XC</th>\n",
|
||
" <th>Child_XC_Unit</th>\n",
|
||
" <th>Child_BIVA_ZRh</th>\n",
|
||
" <th>Child_BIVA_ZXcH</th>\n",
|
||
" <th>Child_PhA</th>\n",
|
||
" <th>Child_PhA_Unit</th>\n",
|
||
" <th>Child_REE_Kcal</th>\n",
|
||
" <th>Child_REE_MJ</th>\n",
|
||
" <th>Child_TEE_Kcal</th>\n",
|
||
" <th>Child_TEE_MJ</th>\n",
|
||
" </tr>\n",
|
||
" </thead>\n",
|
||
" <tbody>\n",
|
||
" <tr>\n",
|
||
" <th>13</th>\n",
|
||
" <td>2025-07-29T18:58:54.0000000Z</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" <td>10000001583275_0055003f5631501320313557</td>\n",
|
||
" <td>KM6479696509</td>\n",
|
||
" <td>Keirstyn</td>\n",
|
||
" <td>Moran</td>\n",
|
||
" <td>1991-02-01T00:00:00.0000000Z</td>\n",
|
||
" <td>34</td>\n",
|
||
" <td>Caucasian</td>\n",
|
||
" <td>Female</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" </tr>\n",
|
||
" </tbody>\n",
|
||
"</table>\n",
|
||
"<p>1 rows × 147 columns</p>\n",
|
||
"</div>"
|
||
],
|
||
"text/plain": [
|
||
" MeasurementDate Comment \\\n",
|
||
"13 2025-07-29T18:58:54.0000000Z NaN \n",
|
||
"\n",
|
||
" ExternalDeviceId ExternalPatientId FirstName \\\n",
|
||
"13 10000001583275_0055003f5631501320313557 KM6479696509 Keirstyn \n",
|
||
"\n",
|
||
" LastName BirthDate Age Ethnicity Gender ... \\\n",
|
||
"13 Moran 1991-02-01T00:00:00.0000000Z 34 Caucasian Female ... \n",
|
||
"\n",
|
||
" Child_XC Child_XC_Unit Child_BIVA_ZRh Child_BIVA_ZXcH Child_PhA \\\n",
|
||
"13 NaN NaN NaN NaN NaN \n",
|
||
"\n",
|
||
" Child_PhA_Unit Child_REE_Kcal Child_REE_MJ Child_TEE_Kcal Child_TEE_MJ \n",
|
||
"13 NaN NaN NaN NaN NaN \n",
|
||
"\n",
|
||
"[1 rows x 147 columns]"
|
||
]
|
||
},
|
||
"execution_count": 24,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"personal_df = pd.read_excel('data/SECA body comp for all patients.xlsx')\n",
|
||
"\n",
|
||
"keirstyn_data = personal_df[personal_df['LastName'].str.contains('Moran', case=False, na=False)]\n",
|
||
"keirstyn_data"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 26,
|
||
"id": "98d9295a",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"VO2 Max: 47.906290322580645\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"v02_max = df['VO2(ml/min)_smoothed'].max()\n",
|
||
"weight = keirstyn_data['Weight'].iloc[0]\n",
|
||
"print(f\"VO2 Max: {v02_max/weight}\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 32,
|
||
"id": "cdfeb309",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"\n",
|
||
"==================================================\n",
|
||
"Optimal Fat Burning Zone (highest fat:carb ratio):\n",
|
||
"Time: 164.0 seconds\n",
|
||
"Fat burn rate: 3.894 kcal/min\n",
|
||
"Carb burn rate: 1.575 kcal/min\n",
|
||
"Fat:Carb ratio: 2.47\n",
|
||
"Heart Rate: 96.7 bpm\n",
|
||
"VO2: 1147.9 ml/min\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"# Find the point where fat burning is highest and carb burning is lowest\n",
|
||
"# Using the smoothed data for more stable results\n",
|
||
"fat_burn_max_idx = df['FAT_smoothed'].idxmax()\n",
|
||
"carb_burn_min_idx = df['CHO_smoothed'].idxmin()\n",
|
||
"\n",
|
||
"# # Get the data at maximum fat burning point\n",
|
||
"# max_fat_row = df.loc[fat_burn_max_idx]\n",
|
||
"# print(f\"Maximum Fat Burning Point:\")\n",
|
||
"# print(f\"Time: {max_fat_row['T(sec)']} seconds\")\n",
|
||
"# print(f\"Fat burn rate: {max_fat_row['FAT_smoothed']:.3f} kcal/min\")\n",
|
||
"# print(f\"Carb burn rate: {max_fat_row['CHO_smoothed']:.3f} kcal/min\")\n",
|
||
"# print(f\"Heart Rate: {max_fat_row['HR(bpm)_smoothed']:.1f} bpm\")\n",
|
||
"# print(f\"VO2: {max_fat_row['VO2(ml/min)_smoothed']:.1f} ml/min\")\n",
|
||
"\n",
|
||
"# print(\"\\n\" + \"=\"*50)\n",
|
||
"\n",
|
||
"# # Get the data at minimum carb burning point\n",
|
||
"# min_carb_row = df.loc[carb_burn_min_idx]\n",
|
||
"# print(f\"Minimum Carbohydrate Burning Point:\")\n",
|
||
"# print(f\"Time: {min_carb_row['T(sec)']} seconds\")\n",
|
||
"# print(f\"Fat burn rate: {min_carb_row['FAT_smoothed']:.3f} kcal/min\")\n",
|
||
"# print(f\"Carb burn rate: {min_carb_row['CHO_smoothed']:.3f} kcal/min\")\n",
|
||
"# print(f\"Heart Rate: {min_carb_row['HR(bpm)_smoothed']:.1f} bpm\")\n",
|
||
"# print(f\"VO2: {min_carb_row['VO2(ml/min)_smoothed']:.1f} ml/min\")\n",
|
||
"\n",
|
||
"print(\"\\n\" + \"=\"*50)\n",
|
||
"\n",
|
||
"# Find the optimal fat burning zone (highest fat:carb ratio)\n",
|
||
"df['fat_carb_ratio'] = df['FAT_smoothed'] / (df['CHO_smoothed'] + 0.00000001) # Add small value to avoid division by zero\n",
|
||
"optimal_fat_idx = df['fat_carb_ratio'].idxmax()\n",
|
||
"optimal_row = df.loc[optimal_fat_idx]\n",
|
||
"\n",
|
||
"print(f\"Optimal Fat Burning Zone (highest fat:carb ratio):\")\n",
|
||
"print(f\"Time: {optimal_row['T(sec)']} seconds\")\n",
|
||
"print(f\"Fat burn rate: {optimal_row['FAT_smoothed']:.3f} kcal/min\")\n",
|
||
"print(f\"Carb burn rate: {optimal_row['CHO_smoothed']:.3f} kcal/min\")\n",
|
||
"print(f\"Fat:Carb ratio: {optimal_row['fat_carb_ratio']:.2f}\")\n",
|
||
"print(f\"Heart Rate: {optimal_row['HR(bpm)_smoothed']:.1f} bpm\")\n",
|
||
"print(f\"VO2: {optimal_row['VO2(ml/min)_smoothed']:.1f} ml/min\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 33,
|
||
"id": "4420cfea",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Found 2 intersections at indices: [18, 47]\n",
|
||
"\n",
|
||
"Last intersection at index 47:\n",
|
||
"Time: 251.0 seconds\n",
|
||
"Fat burn rate: 3.040 kcal/min\n",
|
||
"Carb burn rate: 3.166 kcal/min\n",
|
||
"Heart Rate: 100.5 bpm\n",
|
||
"VO2: 1283.0 ml/min\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"# Find intersections where FAT_smoothed and CHO_smoothed cross each other\n",
|
||
"intersections = []\n",
|
||
"\n",
|
||
"for i in range(1, len(df)):\n",
|
||
" # Check if there's a crossover between consecutive points\n",
|
||
" prev_fat = df.iloc[i-1]['FAT_smoothed']\n",
|
||
" prev_cho = df.iloc[i-1]['CHO_smoothed']\n",
|
||
" curr_fat = df.iloc[i]['FAT_smoothed']\n",
|
||
" curr_cho = df.iloc[i]['CHO_smoothed']\n",
|
||
" \n",
|
||
" # Skip if any values are NaN\n",
|
||
" if pd.isna(prev_fat) or pd.isna(prev_cho) or pd.isna(curr_fat) or pd.isna(curr_cho):\n",
|
||
" continue\n",
|
||
" \n",
|
||
" # Check if lines cross (fat was above/below cho and now it's below/above)\n",
|
||
" if ((prev_fat > prev_cho and curr_fat < curr_cho) or \n",
|
||
" (prev_fat < prev_cho and curr_fat > curr_cho)):\n",
|
||
" intersections.append(i)\n",
|
||
"\n",
|
||
"print(f\"Found {len(intersections)} intersections at indices: {intersections}\")\n",
|
||
"\n",
|
||
"if intersections:\n",
|
||
" # Get the last intersection\n",
|
||
" last_intersection_idx = intersections[-1]\n",
|
||
" last_intersection_row = df.iloc[last_intersection_idx]\n",
|
||
" \n",
|
||
" print(f\"\\nLast intersection at index {last_intersection_idx}:\")\n",
|
||
" print(f\"Time: {last_intersection_row['T(sec)']} seconds\")\n",
|
||
" print(f\"Fat burn rate: {last_intersection_row['FAT_smoothed']:.3f} kcal/min\")\n",
|
||
" print(f\"Carb burn rate: {last_intersection_row['CHO_smoothed']:.3f} kcal/min\")\n",
|
||
" print(f\"Heart Rate: {last_intersection_row['HR(bpm)_smoothed']:.1f} bpm\")\n",
|
||
" print(f\"VO2: {last_intersection_row['VO2(ml/min)_smoothed']:.1f} ml/min\")\n",
|
||
"else:\n",
|
||
" print(\"No intersections found between FAT and CHO curves\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 37,
|
||
"id": "62803668",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"VT1: {'HeartRate': 100.5, 'Speed': 4.0, 'Time': 251.0}\n",
|
||
"VT2: {'HeartRate': 189.71300000000002, 'Speed': 7.5, 'Time': 1524.0}\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"def detect_vt1(df, fat_col=\"FAT_smoothed\", carb_col=\"CHO_smoothed\"):\n",
|
||
" \"\"\"\n",
|
||
" Detect VT1 as the first index where carb burn > fat burn and remains higher.\n",
|
||
" \"\"\"\n",
|
||
" condition = df[carb_col] > df[fat_col]\n",
|
||
" crossover_indices = condition[condition].index\n",
|
||
"\n",
|
||
" if len(crossover_indices) == 0:\n",
|
||
" return None # No crossover found\n",
|
||
" \n",
|
||
" # Find first crossover where carbs remain higher for the rest\n",
|
||
" for idx in crossover_indices:\n",
|
||
" if all(df.loc[idx:][carb_col] > df.loc[idx:][fat_col]):\n",
|
||
" return idx\n",
|
||
" return None\n",
|
||
"\n",
|
||
"\n",
|
||
"def detect_vt2(df, vent_col=\"VE(l/min)_smoothed\", bf_col=\"BF(bpm)_smoothed\", smooth_window=5):\n",
|
||
" \"\"\"\n",
|
||
" Detect VT2 using slope/inflection method.\n",
|
||
" Works with either Ventilation (VE) or Breathing Frequency (Bf).\n",
|
||
" \"\"\"\n",
|
||
" col = vent_col if vent_col in df.columns else bf_col\n",
|
||
" \n",
|
||
" # Use already smoothed data\n",
|
||
" smoothed_col = col\n",
|
||
" \n",
|
||
" # Compute slope (first derivative)\n",
|
||
" df[\"slope\"] = df[smoothed_col].diff()\n",
|
||
" \n",
|
||
" # Detect inflection: largest change in slope (second derivative peak)\n",
|
||
" df[\"second_derivative\"] = df[\"slope\"].diff()\n",
|
||
" inflection_idx = df[\"second_derivative\"].idxmax()\n",
|
||
" \n",
|
||
" return inflection_idx\n",
|
||
"\n",
|
||
"\n",
|
||
"def analyze_thresholds(df_input):\n",
|
||
" # Use the existing dataframe\n",
|
||
" df_copy = df_input.copy()\n",
|
||
" \n",
|
||
" # --- Detect VT1 ---\n",
|
||
" vt1_idx = detect_vt1(df_copy)\n",
|
||
" vt1 = None\n",
|
||
" if vt1_idx is not None:\n",
|
||
" vt1 = {\n",
|
||
" \"HeartRate\": df_copy.loc[vt1_idx, \"HR(bpm)_smoothed\"],\n",
|
||
" \"Speed\": df_copy.loc[vt1_idx, \"Speed\"],\n",
|
||
" \"Time\": df_copy.loc[vt1_idx, \"T(sec)\"]\n",
|
||
" }\n",
|
||
" \n",
|
||
" # --- Detect VT2 ---\n",
|
||
" vt2_idx = detect_vt2(df_copy)\n",
|
||
" vt2 = None\n",
|
||
" if vt2_idx is not None:\n",
|
||
" vt2 = {\n",
|
||
" \"HeartRate\": df_copy.loc[vt2_idx, \"HR(bpm)_smoothed\"],\n",
|
||
" \"Speed\": df_copy.loc[vt2_idx, \"Speed\"],\n",
|
||
" \"Time\": df_copy.loc[vt2_idx, \"T(sec)\"]\n",
|
||
" }\n",
|
||
" \n",
|
||
" return vt1, vt2\n",
|
||
"\n",
|
||
"\n",
|
||
"vt1, vt2 = analyze_thresholds(df)\n",
|
||
"print(\"VT1:\", vt1)\n",
|
||
"print(\"VT2:\", vt2)\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 40,
|
||
"id": "07593b56",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Zone 1 (Active Recovery): 81.7 - 96.7 bpm\n",
|
||
"Zone 2 (Aerobic Base): 96.7 - 100.5 bpm\n",
|
||
"Zone 3 (Aerobic): 100.5 - 179.7 bpm\n",
|
||
"Zone 4 (Lactate Threshold): 179.7 - 199.7 bpm\n",
|
||
"Zone 5 (VO2 Max): 199.7+ bpm\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"zone_1_start = optimal_row['HR(bpm)_smoothed'] - 15\n",
|
||
"zone_2_start = optimal_row['HR(bpm)_smoothed']\n",
|
||
"zone_3_start = vt1\n",
|
||
"zone_4_start = vt2['HeartRate'] - 10\n",
|
||
"zone_5_start = vt2['HeartRate'] + 10\n",
|
||
"\n",
|
||
"zone_1_end = zone_2_start\n",
|
||
"zone_2_end = vt1['HeartRate']\n",
|
||
"zone_3_end = zone_4_start\n",
|
||
"zone_4_end = zone_5_start\n",
|
||
"\n",
|
||
"print(f\"Zone 1 (Active Recovery): {zone_1_start:.1f} - {zone_1_end:.1f} bpm\")\n",
|
||
"print(f\"Zone 2 (Aerobic Base): {zone_2_start:.1f} - {zone_2_end:.1f} bpm\")\n",
|
||
"print(f\"Zone 3 (Aerobic): {zone_3_start['HeartRate']:.1f} - {zone_3_end:.1f} bpm\")\n",
|
||
"print(f\"Zone 4 (Lactate Threshold): {zone_4_start:.1f} - {zone_4_end:.1f} bpm\")\n",
|
||
"print(f\"Zone 5 (VO2 Max): {zone_5_start:.1f}+ bpm\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 60,
|
||
"id": "c90415b2",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"VO2 Max detected at index 202:\n",
|
||
"Time: 985.0 seconds\n",
|
||
"VO2 Breath: 58.2 ml/breath\n",
|
||
"VO2: 2167.8 ml/min\n",
|
||
"VO2 per kg: 38.8 ml/kg/min\n",
|
||
"Heart Rate: 170.5 bpm\n",
|
||
"Speed: 6.0 km/h\n",
|
||
"VO2 Breath Slope: -0.02\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"# Calculate the slope of VO2 Breath (first derivative)\n",
|
||
"df['vo2_breath_slope'] = df['VO2 Breath_smoothed'].diff()\n",
|
||
"\n",
|
||
"# Find points where slope is consistently zero or negative\n",
|
||
"# We'll use a rolling window to check for consistent negative/zero slope\n",
|
||
"window = len(df) // 3 # Number of consecutive points to check\n",
|
||
"\n",
|
||
"# Calculate rolling mean of slope to smooth out noise\n",
|
||
"df['vo2_breath_slope_smoothed'] = df['vo2_breath_slope'].rolling(window=window).mean()\n",
|
||
"\n",
|
||
"# Find where slope becomes consistently zero or negative\n",
|
||
"mask = df['vo2_breath_slope_smoothed'] <= 0\n",
|
||
"consistent_negative_indices = mask[mask].index\n",
|
||
"\n",
|
||
"if len(consistent_negative_indices) > 0:\n",
|
||
" # Find the first point where slope becomes consistently negative/zero\n",
|
||
" vo2_max_idx = consistent_negative_indices[0]\n",
|
||
" vo2_max_row = df.loc[vo2_max_idx]\n",
|
||
" \n",
|
||
" print(f\"VO2 Max detected at index {vo2_max_idx}:\")\n",
|
||
" print(f\"Time: {vo2_max_row['T(sec)']} seconds\")\n",
|
||
" print(f\"VO2 Breath: {vo2_max_row['VO2 Breath_smoothed']:.1f} ml/breath\")\n",
|
||
" print(f\"VO2: {vo2_max_row['VO2(ml/min)_smoothed']:.1f} ml/min\")\n",
|
||
" print(f\"VO2 per kg: {vo2_max_row['VO2(ml/min)_smoothed']/weight:.1f} ml/kg/min\")\n",
|
||
" print(f\"Heart Rate: {vo2_max_row['HR(bpm)_smoothed']:.1f} bpm\")\n",
|
||
" print(f\"Speed: {vo2_max_row['Speed']} km/h\")\n",
|
||
" print(f\"VO2 Breath Slope: {vo2_max_row['vo2_breath_slope_smoothed']:.2f}\")\n",
|
||
"else:\n",
|
||
" # If no consistent negative slope found, use the maximum VO2 Breath value\n",
|
||
" vo2_max_idx = df['VO2 Breath_smoothed'].idxmax()\n",
|
||
" vo2_max_row = df.loc[vo2_max_idx]\n",
|
||
" \n",
|
||
" print(f\"No consistent negative slope found. Using peak VO2 Breath at index {vo2_max_idx}:\")\n",
|
||
" print(f\"Time: {vo2_max_row['T(sec)']} seconds\")\n",
|
||
" print(f\"VO2 Breath: {vo2_max_row['VO2 Breath_smoothed']:.1f} ml/breath\")\n",
|
||
" print(f\"VO2: {vo2_max_row['VO2(ml/min)_smoothed']:.1f} ml/min\")\n",
|
||
" print(f\"VO2 per kg: {vo2_max_row['VO2(ml/min)_smoothed']/weight:.1f} ml/kg/min\")\n",
|
||
" print(f\"Heart Rate: {vo2_max_row['HR(bpm)_smoothed']:.1f} bpm\")\n",
|
||
" print(f\"Speed: {vo2_max_row['Speed']} km/h\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 66,
|
||
"id": "c3b2cc59",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"VO2 Pulse and HR slopes diverge consistently starting at index 89:\n",
|
||
"Time: 485.0 seconds\n",
|
||
"VO2 Pulse (smoothed): 13.91\n",
|
||
"Heart Rate (smoothed): 136.2 bpm\n",
|
||
"VO2 Pulse Slope: 0.672\n",
|
||
"HR Slope: 1.000\n",
|
||
"Slope Difference: 1.006\n",
|
||
"VO2: 1897.8 ml/min\n",
|
||
"Speed: 4.5 km/h\n",
|
||
"Threshold used: 0.615\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"# Calculate slopes for both VO2 Pulse and HR\n",
|
||
"df['vo2_pulse_slope'] = df['VO2 Pulse_smoothed'].diff()\n",
|
||
"df['hr_slope'] = df['HR(bpm)_smoothed'].diff()\n",
|
||
"\n",
|
||
"# Calculate the difference between the slopes\n",
|
||
"df['slope_difference'] = abs(df['vo2_pulse_slope'] - df['hr_slope'])\n",
|
||
"\n",
|
||
"# Find where the slope difference becomes consistently large (slopes diverge)\n",
|
||
"# Use a rolling window to smooth out noise\n",
|
||
"window_size = len(df) // 5 # Adjust window size as needed\n",
|
||
"df['slope_difference_smoothed'] = df['slope_difference'].rolling(window=window_size).mean()\n",
|
||
"\n",
|
||
"# Find the threshold - we'll use the 75th percentile of slope differences as threshold\n",
|
||
"threshold = df['slope_difference_smoothed'].quantile(0.75)\n",
|
||
"\n",
|
||
"# Find points where slope difference exceeds threshold\n",
|
||
"divergence_mask = df['slope_difference_smoothed'] > threshold\n",
|
||
"divergence_indices = divergence_mask[divergence_mask].index\n",
|
||
"\n",
|
||
"if len(divergence_indices) > 0:\n",
|
||
" # Find the first sustained divergence point\n",
|
||
" min_consecutive_points = 5\n",
|
||
" consistent_divergence_idx = None\n",
|
||
" \n",
|
||
" for start_idx in divergence_indices:\n",
|
||
" # Check if divergence is sustained for consecutive points\n",
|
||
" consecutive_count = 0\n",
|
||
" for j in range(start_idx, min(start_idx + min_consecutive_points, len(df))):\n",
|
||
" if j in divergence_indices:\n",
|
||
" consecutive_count += 1\n",
|
||
" else:\n",
|
||
" break\n",
|
||
" \n",
|
||
" if consecutive_count >= min_consecutive_points:\n",
|
||
" consistent_divergence_idx = start_idx\n",
|
||
" break\n",
|
||
" \n",
|
||
" if consistent_divergence_idx is not None:\n",
|
||
" divergence_row = df.iloc[consistent_divergence_idx]\n",
|
||
" \n",
|
||
" print(f\"VO2 Pulse and HR slopes diverge consistently starting at index {consistent_divergence_idx}:\")\n",
|
||
" print(f\"Time: {divergence_row['T(sec)']} seconds\")\n",
|
||
" print(f\"VO2 Pulse (smoothed): {divergence_row['VO2 Pulse_smoothed']:.2f}\")\n",
|
||
" print(f\"Heart Rate (smoothed): {divergence_row['HR(bpm)_smoothed']:.1f} bpm\")\n",
|
||
" print(f\"VO2 Pulse Slope: {divergence_row['vo2_pulse_slope']:.3f}\")\n",
|
||
" print(f\"HR Slope: {divergence_row['hr_slope']:.3f}\")\n",
|
||
" print(f\"Slope Difference: {divergence_row['slope_difference_smoothed']:.3f}\")\n",
|
||
" print(f\"VO2: {divergence_row['VO2(ml/min)_smoothed']:.1f} ml/min\")\n",
|
||
" print(f\"Speed: {divergence_row['Speed']} km/h\")\n",
|
||
" print(f\"Threshold used: {threshold:.3f}\")\n",
|
||
" else:\n",
|
||
" print(f\"No sustained divergence found. Threshold: {threshold:.3f}\")\n",
|
||
" # Show the point with maximum slope difference instead\n",
|
||
" max_diff_idx = df['slope_difference_smoothed'].idxmax()\n",
|
||
" max_diff_row = df.iloc[max_diff_idx]\n",
|
||
" \n",
|
||
" print(f\"\\nPoint with maximum slope difference at index {max_diff_idx}:\")\n",
|
||
" print(f\"Time: {max_diff_row['T(sec)']} seconds\")\n",
|
||
" print(f\"VO2 Pulse (smoothed): {max_diff_row['VO2 Pulse_smoothed']:.2f}\")\n",
|
||
" print(f\"Heart Rate (smoothed): {max_diff_row['HR(bpm)_smoothed']:.1f} bpm\")\n",
|
||
" print(f\"Slope Difference: {max_diff_row['slope_difference_smoothed']:.3f}\")\n",
|
||
"else:\n",
|
||
" print(\"No significant slope divergence found between VO2 Pulse and HR\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "672d68f3",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Maximum FAT_smoothed occurs at index 30:\n",
|
||
"Heart Rate (smoothed): 96.7 bpm\n",
|
||
"FAT (smoothed): 3.894 kcal/min\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"max_fat_smoothed_idx = df['FAT_smoothed'].idxmax()\n",
|
||
"max_fat_smoothed_row = df.loc[max_fat_smoothed_idx]\n",
|
||
"max_heart_rate = 220 - keirstyn_data['Age'].iloc[0]\n",
|
||
"\n",
|
||
"print(f\"Maximum FAT_smoothed occurs at index {max_fat_smoothed_idx}:\")\n",
|
||
"print(f\"Heart Rate (smoothed): {max_fat_smoothed_row['HR(bpm)_smoothed']:.1f} bpm\")\n",
|
||
"print(f\"FAT (smoothed): {max_fat_smoothed_row['FAT_smoothed']:.3f} kcal/min\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "fe3b7605",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": []
|
||
}
|
||
],
|
||
"metadata": {
|
||
"kernelspec": {
|
||
"display_name": "report_generation",
|
||
"language": "python",
|
||
"name": "python3"
|
||
},
|
||
"language_info": {
|
||
"codemirror_mode": {
|
||
"name": "ipython",
|
||
"version": 3
|
||
},
|
||
"file_extension": ".py",
|
||
"mimetype": "text/x-python",
|
||
"name": "python",
|
||
"nbconvert_exporter": "python",
|
||
"pygments_lexer": "ipython3",
|
||
"version": "3.12.3"
|
||
}
|
||
},
|
||
"nbformat": 4,
|
||
"nbformat_minor": 5
|
||
}
|