From 7669f2ac4bd4b61b57dd33a900c80b558a20553d Mon Sep 17 00:00:00 2001
From: MuslemRahimi <moslem_rahimi@hotmail.de>
Date: Tue, 27 Aug 2024 10:34:51 +0200
Subject: [PATCH] update cron job

---
 app/cron_var.py | 67 ++++++++++++++++++++++++++-----------------------
 1 file changed, 36 insertions(+), 31 deletions(-)

diff --git a/app/cron_var.py b/app/cron_var.py
index fcac73a..2f49348 100755
--- a/app/cron_var.py
+++ b/app/cron_var.py
@@ -4,9 +4,11 @@ import numpy as np
 import ujson
 import asyncio
 import sqlite3
+import os
 from tqdm import tqdm
 
 async def save_json(symbol, data):
+    os.makedirs("json/var", exist_ok=True)  # Ensure directory exists
     with open(f"json/var/{symbol}.json", 'w') as file:
         ujson.dump(data, file)
 
@@ -39,10 +41,10 @@ def compute_var(df):
     df = df.dropna()
     # Calculate VaR at 95% confidence level
     confidence_level = 0.95
-    var = abs(np.percentile(df['Returns'], 100 * (1 - confidence_level)))
-    var_N_days = round(var * np.sqrt(5)*100,2) # N days
+    var = np.percentile(df['Returns'], 100 * (1 - confidence_level))
+    var_N_days = round(var * np.sqrt(len(df)) * 100, 2)  # N days: the length of df represents the N days
 
-    return -var_N_days #{'rating': risk_rating, 'var': -var_N_days, 'outlook': outlook}
+    return var_N_days  # Positive value represents a loss
 
 async def run():
     start_date = "2015-01-01"
@@ -70,36 +72,39 @@ async def run():
     total_symbols = stocks_symbols + etf_symbols + crypto_symbols
 
     for symbol in tqdm(total_symbols):
-        if symbol in etf_symbols:  
-            query_con = etf_con
-        elif symbol in crypto_symbols:  
-            query_con = crypto_con
-        elif symbol in stocks_symbols:  
-            query_con = con
-
-        query_template = """
-                SELECT
-                    date, open, high, low, close, volume
-                FROM
-                    "{symbol}"
-                WHERE
-                    date BETWEEN ? AND ?
-            """
-        query = query_template.format(symbol=symbol)
-        df = pd.read_sql_query(query, query_con, params=(start_date, end_date))
-
-        # Convert date to datetime
-        df['date'] = pd.to_datetime(df['date'])
-
-        # Group by year and month
-        monthly_groups = df.groupby(df['date'].dt.to_period('M'))
-
-        history = []
         try:
+            if symbol in etf_symbols:  
+                query_con = etf_con
+            elif symbol in crypto_symbols:  
+                query_con = crypto_con
+            elif symbol in stocks_symbols:  
+                query_con = con
+            else:
+                continue
+
+            query_template = """
+                    SELECT
+                        date, open, high, low, close, volume
+                    FROM
+                        "{symbol}"
+                    WHERE
+                        date BETWEEN ? AND ?
+                """
+            query = query_template.format(symbol=symbol)
+            df = pd.read_sql_query(query, query_con, params=(start_date, end_date))
+
+            # Convert date to datetime
+            df['date'] = pd.to_datetime(df['date'])
+
+            # Group by year and month
+            monthly_groups = df.groupby(df['date'].dt.to_period('M'))
+            history = []
+
             for period, group in monthly_groups:
-                var_data = compute_var(group)
-                history.append({'date': str(period), 'var': var_data})
-            
+                if len(group) >=19:  # Check if the month has at least 19 data points
+                    var_data = compute_var(group)
+                    history.append({'date': str(period), 'var': var_data})
+
             risk_rating = assign_risk_rating(abs(history[-1]['var']))
             outlook = 'Neutral'
             if risk_rating < 5: