Spaces:
Running
Running
File size: 11,342 Bytes
022ace2 7ca4926 c6aed6b 022ace2 c6aed6b 022ace2 c6aed6b 022ace2 7ca4926 022ace2 c6aed6b 7ca4926 c6aed6b 7ca4926 c6aed6b 7ca4926 c6aed6b 7ca4926 022ace2 c6aed6b 022ace2 7ca4926 c46f583 7ca4926 022ace2 c7eeac3 7ca4926 c46f583 7ca4926 c7eeac3 022ace2 c46f583 c6aed6b 7ca4926 c6aed6b c46f583 c6aed6b c46f583 c6aed6b c46f583 c6aed6b c46f583 c6aed6b c46f583 022ace2 c6aed6b 7ca4926 c46f583 7ca4926 c46f583 7ca4926 c46f583 7ca4926 c46f583 7ca4926 c46f583 7ca4926 c46f583 7ca4926 c46f583 7ca4926 c46f583 7ca4926 c46f583 7ca4926 c46f583 7ca4926 c46f583 7ca4926 c6aed6b 022ace2 c46f583 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 |
import gradio as gr
import pandas as pd
import time
import traceback
from OpenAITools.FetchTools import fetch_clinical_trials
from langchain_openai import ChatOpenAI
from langchain_groq import ChatGroq
from OpenAITools.CrinicalTrialTools import SimpleClinicalTrialAgent, GraderAgent, LLMTranslator, generate_ex_question_English
# モデルとエージェントの初期化
groq = ChatGroq(model_name="llama3-70b-8192", temperature=0)
translator = LLMTranslator(groq)
CriteriaCheckAgent = SimpleClinicalTrialAgent(groq)
grader_agent = GraderAgent(groq)
# エラーハンドリング付きでエージェント評価を実行する関数
def evaluate_with_retry(agent, criteria, question, max_retries=3):
"""エラーハンドリング付きでエージェント評価を実行"""
for attempt in range(max_retries):
try:
return agent.evaluate_eligibility(criteria, question)
except Exception as e:
if "missing variables" in str(e):
# プロンプトテンプレートの変数エラーの場合
print(f"プロンプトテンプレートエラー (試行 {attempt + 1}/{max_retries}): {e}")
return "評価エラー: プロンプトテンプレートの設定に問題があります"
elif "no healthy upstream" in str(e) or "InternalServerError" in str(e):
# Groqサーバーエラーの場合
print(f"Groqサーバーエラー (試行 {attempt + 1}/{max_retries}): {e}")
if attempt < max_retries - 1:
time.sleep(2) # 2秒待機してリトライ
continue
else:
return "評価エラー: サーバーに接続できませんでした"
else:
print(f"予期しないエラー (試行 {attempt + 1}/{max_retries}): {e}")
if attempt < max_retries - 1:
time.sleep(1)
continue
else:
return f"評価エラー: {str(e)}"
return "評価エラー: 最大リトライ回数に達しました"
def evaluate_grade_with_retry(agent, judgment, max_retries=3):
"""エラーハンドリング付きでグレード評価を実行"""
for attempt in range(max_retries):
try:
return agent.evaluate_eligibility(judgment)
except Exception as e:
if "no healthy upstream" in str(e) or "InternalServerError" in str(e):
print(f"Groqサーバーエラー (グレード評価 - 試行 {attempt + 1}/{max_retries}): {e}")
if attempt < max_retries - 1:
time.sleep(2)
continue
else:
return "unclear"
else:
print(f"予期しないエラー (グレード評価 - 試行 {attempt + 1}/{max_retries}): {e}")
if attempt < max_retries - 1:
time.sleep(1)
continue
else:
return "unclear"
return "unclear"
# データフレームを生成する関数
def generate_dataframe(age, sex, tumor_type, GeneMutation, Meseable, Biopsiable):
try:
# 入力検証
if not all([age, sex, tumor_type]):
return pd.DataFrame(), pd.DataFrame()
# 日本語の腫瘍タイプを英語に翻訳
try:
TumorName = translator.translate(tumor_type)
except Exception as e:
print(f"翻訳エラー: {e}")
TumorName = tumor_type # 翻訳に失敗した場合は元の値を使用
# 質問文を生成
try:
ex_question = generate_ex_question_English(age, sex, TumorName, GeneMutation, Meseable, Biopsiable)
except Exception as e:
print(f"質問生成エラー: {e}")
return pd.DataFrame(), pd.DataFrame()
# 臨床試験データの取得
try:
df = fetch_clinical_trials(TumorName)
if df.empty:
print("臨床試験データが見つかりませんでした")
return pd.DataFrame(), pd.DataFrame()
except Exception as e:
print(f"臨床試験データ取得エラー: {e}")
return pd.DataFrame(), pd.DataFrame()
df['AgentJudgment'] = None
df['AgentGrade'] = None
# 臨床試験の適格性の評価
NCTIDs = list(df['NCTID'])
progress = gr.Progress(track_tqdm=True)
for i, nct_id in enumerate(NCTIDs):
try:
target_criteria = df.loc[df['NCTID'] == nct_id, 'Eligibility Criteria'].values[0]
# エラーハンドリング付きで評価実行
agent_judgment = evaluate_with_retry(CriteriaCheckAgent, target_criteria, ex_question)
agent_grade = evaluate_grade_with_retry(grader_agent, agent_judgment)
# データフレームの更新
df.loc[df['NCTID'] == nct_id, 'AgentJudgment'] = agent_judgment
df.loc[df['NCTID'] == nct_id, 'AgentGrade'] = agent_grade
except Exception as e:
print(f"NCTID {nct_id} の評価中にエラー: {e}")
df.loc[df['NCTID'] == nct_id, 'AgentJudgment'] = f"エラー: {str(e)}"
df.loc[df['NCTID'] == nct_id, 'AgentGrade'] = "unclear"
progress((i + 1) / len(NCTIDs))
# 列を指定した順に並び替え
columns_order = ['NCTID', 'AgentGrade', 'Title', 'AgentJudgment', 'Japanes Locations',
'Primary Completion Date', 'Cancer', 'Summary', 'Eligibility Criteria']
# 存在する列のみを選択
available_columns = [col for col in columns_order if col in df.columns]
df = df[available_columns]
return df, df # フィルタ用と表示用にデータフレームを返す
except Exception as e:
print(f"データフレーム生成中に予期しないエラー: {e}")
traceback.print_exc()
return pd.DataFrame(), pd.DataFrame()
# CSVとして保存しダウンロードする関数
def download_filtered_csv(df):
try:
if df is None or len(df) == 0:
return None
file_path = "filtered_data.csv"
df.to_csv(file_path, index=False)
return file_path
except Exception as e:
print(f"CSV保存エラー: {e}")
return None
# 全体結果をCSVとして保存しダウンロードする関数
def download_full_csv(df):
try:
if df is None or len(df) == 0:
return None
file_path = "full_data.csv"
df.to_csv(file_path, index=False)
return file_path
except Exception as e:
print(f"CSV保存エラー: {e}")
return None
# Gradioインターフェースの作成
with gr.Blocks(title="臨床試験適格性評価", theme=gr.themes.Soft()) as demo:
gr.Markdown("## 臨床試験適格性評価インターフェース")
gr.Markdown("⚠️ **注意**: サーバーエラーが発生する場合があります。エラーが続く場合は少し時間をおいてから再試行してください。")
# 各種入力フィールド
with gr.Row():
with gr.Column():
age_input = gr.Textbox(label="Age", placeholder="例: 65", value="")
sex_input = gr.Dropdown(choices=["男性", "女性"], label="Sex", value=None)
tumor_type_input = gr.Textbox(label="Tumor Type", placeholder="例: gastric cancer", value="")
with gr.Column():
gene_mutation_input = gr.Textbox(label="Gene Mutation", placeholder="例: HER2", value="")
measurable_input = gr.Dropdown(choices=["有り", "無し", "不明"], label="Measurable Tumor", value=None)
biopsiable_input = gr.Dropdown(choices=["有り", "無し", "不明"], label="Biopsiable Tumor", value=None)
# データフレーム表示エリア
dataframe_output = gr.DataFrame(
headers=["NCTID", "AgentGrade", "Title", "AgentJudgment", "Status"],
datatype=["str", "str", "str", "str", "str"],
value=None
)
# 内部状態用の非表示コンポーネント
original_df_state = gr.State(value=None)
filtered_df_state = gr.State(value=None)
# ボタン類
with gr.Row():
generate_button = gr.Button("Generate Clinical Trials Data", variant="primary")
with gr.Row():
yes_button = gr.Button("Show Eligible Trials", variant="secondary")
no_button = gr.Button("Show Ineligible Trials", variant="secondary")
unclear_button = gr.Button("Show Unclear Trials", variant="secondary")
with gr.Row():
download_filtered_button = gr.Button("Download Filtered Data")
download_full_button = gr.Button("Download Full Data")
# ダウンロードファイル
download_filtered_output = gr.File(label="Download Filtered Data", visible=False)
download_full_output = gr.File(label="Download Full Data", visible=False)
# イベントハンドリング
def update_dataframe_and_state(age, sex, tumor_type, gene_mutation, measurable, biopsiable):
"""データフレーム生成と状態更新"""
df, _ = generate_dataframe(age, sex, tumor_type, gene_mutation, measurable, biopsiable)
return df, df, df
def filter_and_update(original_df, grade):
"""フィルタリングと表示更新"""
if original_df is None or len(original_df) == 0:
return original_df, original_df
try:
df_filtered = original_df[original_df['AgentGrade'] == grade]
return df_filtered, df_filtered
except Exception as e:
print(f"フィルタリングエラー: {e}")
return original_df, original_df
# ボタン動作の設定
generate_button.click(
fn=update_dataframe_and_state,
inputs=[age_input, sex_input, tumor_type_input, gene_mutation_input, measurable_input, biopsiable_input],
outputs=[dataframe_output, original_df_state, filtered_df_state]
)
yes_button.click(
fn=lambda df: filter_and_update(df, "yes"),
inputs=[original_df_state],
outputs=[dataframe_output, filtered_df_state]
)
no_button.click(
fn=lambda df: filter_and_update(df, "no"),
inputs=[original_df_state],
outputs=[dataframe_output, filtered_df_state]
)
unclear_button.click(
fn=lambda df: filter_and_update(df, "unclear"),
inputs=[original_df_state],
outputs=[dataframe_output, filtered_df_state]
)
download_filtered_button.click(
fn=download_filtered_csv,
inputs=[filtered_df_state],
outputs=[download_filtered_output]
)
download_full_button.click(
fn=download_full_csv,
inputs=[original_df_state],
outputs=[download_full_output]
)
if __name__ == "__main__":
demo.launch(
server_name="0.0.0.0",
server_port=7860,
share=False,
debug=False,
show_error=True
) |