From c285883d71ef8a362b012983dadc7ce4256b40f6 Mon Sep 17 00:00:00 2001
From: Administrator <admin@example.com>
Date: 星期五, 23 五月 2025 01:52:07 +0800
Subject: [PATCH] bug修复
---
l2_data_parser.py | 137 +++++++++++++++++++++++++++++++++++++++++++--
1 files changed, 130 insertions(+), 7 deletions(-)
diff --git a/l2_data_parser.py b/l2_data_parser.py
index abdbe66..e96b4ed 100644
--- a/l2_data_parser.py
+++ b/l2_data_parser.py
@@ -10,6 +10,7 @@
from db import mysql_data_delegate as mysql_data
from huaxin_client.l2_client_test import L2TransactionDataManager
from log_module import log_export
+from utils import tool
def __get_target_codes(day):
@@ -50,13 +51,6 @@
fdatas = []
df = pd.read_csv(f"{base_path}/Transaction.csv")
category_revenue = df.groupby('BuyNo')['TradeVolume'].sum()
- total_count = 0
- for chunk in chunks:
- total_count += len(chunk)
- for chunk in chunks:
- total_count += len(chunk)
-
-
with open(f"{base_path}/Transaction.csv", 'r', encoding='utf-8') as file:
csv_reader = csv.reader(file)
@@ -231,6 +225,131 @@
writer.writerow(row)
+def pre_process_transactions(csv_path="E:/娴嬭瘯鏁版嵁/Transaction_Test.csv"):
+ def str_to_float(s):
+ try:
+ # 绉婚櫎鍗曚綅骞惰浆鎹�
+ return round(float(s.split("@")[0]), 2)
+ except:
+ return float("nan")
+
+ def code_format(s):
+ try:
+ code = "{0:0>6}".format(s)
+ return code
+ except:
+ return ''
+
+ # [ExchangeID, SecurityID, TradeTime, TradePrice, TradeVolume, ExecType, MainSeq, SubSeq, BuyNo, SellNo, Info1, Info2,
+ # Info3, TradeBSFlag, BizIndex, LocalTimeStamp]
+ # transaction_data = {
+ # "SecurityID": ['300920', '300920', '300920', '300920'],
+ # "TradeTime": [91500040, 91500041, 91500042, 92000040],
+ # "TradePrice": [15.0, 16.2, 15.2, 16.3],
+ # "TradeVolume": [100, 100, 200, 300],
+ # "BuyNo": [0, 1, 1, 1]
+ # }
+ # 瀹氫箟鑱氬悎鍑芥暟
+ def first_last(group):
+ return pd.Series({
+ 'TotalAmount': group['TradeAmount'].sum(),
+ 'TotalVolume': group['TradeVolume'].sum(),
+ 'StartTime': group['TradeTime'].iloc[0],
+ 'StartPrice': group['TradePrice'].iloc[0],
+ 'EndTime': group['TradeTime'].iloc[-1],
+ 'EndPrice': group['TradePrice'].iloc[-1]
+ })
+
+ dtype = {
+ 'SecurityID': 'category', # 浣庡熀鏁板垎绫绘暟鎹�
+ }
+ chunk_size = 100000
+ # 鍒涘缓DataFrame
+ chunks = pd.read_csv(csv_path, chunksize=chunk_size)
+ result_list = []
+ index = 0
+ for df in chunks:
+ index += 1
+ child_path = csv_path.replace(".csv", f"_{index}.csv")
+ if os.path.exists(child_path):
+ continue
+ print(f"澶勭悊绗瑊index}鎵规")
+ df["TradePrice"] = df["TradePrice"].apply(str_to_float)
+ df["SecurityID"] = df["SecurityID"].apply(code_format)
+ df = df[df["SecurityID"].str.startswith(("30", "00", "60"), na=False)]
+ # 璁＄畻鎴愪氦閲戦
+ df['TradeAmount'] = df['TradePrice'] * df['TradeVolume']
+
+
+ # 鎸塖ecurityID鍜孊uyNo鍒嗙粍
+ grouped = df.groupby(['SecurityID', 'BuyNo'])
+
+ # 搴旂敤鑱氬悎鍑芥暟
+ chunk_result = grouped.apply(first_last).reset_index()
+
+ chunk_result.to_csv(child_path, index=False)
+ print(f"澶勭悊瀹屾瘯锛屾�诲叡{index}鎵�")
+
+
+def pre_process_ngtstick(csv_path="E:/娴嬭瘯鏁版嵁/NGTSTick_Test.csv"):
+ def str_to_float(s):
+ try:
+ # 绉婚櫎鍗曚綅骞惰浆鎹�
+ return round(float(s.split("@")[0]), 2)
+ except:
+ return float("nan")
+
+ def code_format(s):
+ try:
+ code = "{0:0>6}".format(s)
+ return code
+ except:
+ return ''
+
+ # 瀹氫箟鑱氬悎鍑芥暟
+ def first_last(group):
+ return pd.Series({
+ 'TotalAmount': group['TradeMoney'].sum(),
+ 'TotalVolume': group['Volume'].sum(),
+ 'StartTime': group['TickTime'].iloc[0],
+ 'StartPrice': group['Price'].iloc[0],
+ 'EndTime': group['TickTime'].iloc[-1],
+ 'EndPrice': group['Price'].iloc[-1]
+ })
+
+ # [ExchangeID,SecurityID,MainSeq,SubSeq,TickTime,TickType,BuyNo,SellNo,Price,Volume,TradeMoney,Side,TradeBSFlag,MDSecurityStat,Info1,Info2,Info3,LocalTimeStamp]
+
+ chunk_size = 200000
+ # 鍒涘缓DataFrame
+ chunks = pd.read_csv(csv_path, chunksize=chunk_size)
+ result_list = []
+ index = 0
+ for df in chunks:
+ index += 1
+ child_path = csv_path.replace(".csv", f"_{index}.csv")
+ if os.path.exists(child_path):
+ continue
+ print(f"澶勭悊绗瑊index}鎵规")
+ df = df[df["TickType"] == 'T']
+ df["Price"] = df["Price"].apply(str_to_float)
+ df["SecurityID"] = df["SecurityID"].apply(code_format)
+
+ df = df[df["SecurityID"].str.startswith(("30", "00", "60"), na=False)]
+
+ # 璁＄畻鎴愪氦閲戦
+ df['TradeMoney'] = df["TradeMoney"].apply(str_to_float)
+ # 鎸塖ecurityID鍜孊uyNo鍒嗙粍
+ grouped = df.groupby(['SecurityID', 'BuyNo'])
+ # 搴旂敤鑱氬悎鍑芥暟
+ chunk_result = grouped.apply(first_last).reset_index()
+ chunk_result.to_csv(child_path, index=False)
+ print(f"澶勭悊瀹屾瘯锛屾�诲叡{index}鎵�")
+
+
+if __name__ == '__main__1':
+ # df = pd.read_csv(f"E:/娴嬭瘯鏁版嵁/Transaction_Test.csv")
+ pre_process_ngtstick()
+
# 鍛戒护妯″紡 /home/userzjj/app/gp-server/l2_data_parser Transaction 2025-05-08
# 瑙ｆ瀽澶у崟锛� /home/userzjj/app/gp-server/l2_data_parser ExtractDealBigOrder 2025-05-09 /home/userzjj/鏈�缁堟垚浜ゆ暟鎹�20250509.txt 000555
if __name__ == '__main__':
@@ -251,6 +370,10 @@
parse_ngtstick(day)
elif _type == 'MarketData':
parse_market_data(day)
+ elif _type == 'Transaction_New':
+ pre_process_transactions(f"/home/userzjj/ftp/{day}/Transaction.csv")
+ elif _type == 'NGTSTick_New':
+ pre_process_ngtstick(f"/home/userzjj/ftp/{day}/NGTSTick.csv")
elif _type == 'ExtractDealBigOrder':
# 鎻愬彇鎵�鏈夋垚浜ょ殑澶у崟
if len(params) > 2:
--
Gitblit v1.8.0