Getting Historical Data of Reliance In Pandas Using Python and Zerodha API
Now, coming back to our old discussion of getting the historical data. So lets backtest from the start of 1/1/2019 to today. Although Kite gives data from far behind, Let’s just hold onto last few years to get faster results. You may chose older dates as you may please.
import datetime
starting_date=datetime.datetime(2019, 1, 1)
ending_date=datetime.datetime.today()
The starting_date and ending_date variables are now initiated with two datetime datatypes.
Now, lets fetch the data –
zap=kite.historical_data(token,starting_date,ending_date,"day")
zap= pd.DataFrame(zap)
print(zap.head(10))
It will output –
The initial display exhibits the first 10 rows, utilizing the ‘head’ function to offer a more concise data view of the Pandas Dataframe. It effectively presents OHLC data and volume for each day.
date open high low close volume
0 2019-01-01 00:00:00+05:30 1072.60 1074.55 1058.15 1068.55 4674622
1 2019-01-02 00:00:00+05:30 1062.35 1074.25 1049.45 1054.60 7495772
2 2019-01-03 00:00:00+05:30 1055.65 1062.45 1039.05 1041.60 7812061
3 2019-01-04 00:00:00+05:30 1046.05 1052.75 1030.50 1047.25 8880761
4 2019-01-07 00:00:00+05:30 1055.20 1066.10 1049.45 1053.05 5784262
5 2019-01-08 00:00:00+05:30 1053.35 1058.00 1044.70 1052.95 5901336
6 2019-01-09 00:00:00+05:30 1059.95 1064.70 1047.30 1058.75 6049942
7 2019-01-10 00:00:00+05:30 1055.90 1059.00 1051.40 1055.65 4280616
8 2019-01-11 00:00:00+05:30 1055.75 1061.65 1037.65 1046.65 6781266
9 2019-01-14 00:00:00+05:30 1043.75 1049.00 1035.55 1045.45 4313661
Now, Let’s try to find data for each minute –
zap=kite.historical_data(token,starting_date,ending_date,"minute")
zap= pd.DataFrame(zap)
print(zap.head(10))
It throws an error –
---------------------------------------------------------------------------
InputException Traceback (most recent call last)
Cell In[6], line 1
----> 1 zap=kite.historical_data(token,starting_date,ending_date,"minute")
2 zap= pd.DataFrame(zap)
3 print(zap.head(10))
File ~/apps/zerodha/../kiteconnect/connect.py:503, in KiteConnect.historical_data(self, instrument_token, from_date, to_date, interval, continuous, oi)
500 from_date_string = from_date.strftime(date_string_format) if type(from_date) == datetime.datetime else from_date
501 to_date_string = to_date.strftime(date_string_format) if type(to_date) == datetime.datetime else to_date
--> 503 data = self._get("market.historical", {
504 "instrument_token": instrument_token,
505 "from": from_date_string,
506 "to": to_date_string,
507 "interval": interval,
508 "continuous": 1 if continuous else 0,
509 "oi": 1 if oi else 0
510 })
512 return self._format_historical(data, oi)
File ~/apps/zerodha/../kiteconnect/connect.py:601, in KiteConnect._get(self, route, params)
599 def _get(self, route, params=None):
600 """Alias for sending a GET request."""
--> 601 return self._request(route, "GET", params)
File ~/apps/zerodha/../kiteconnect/connect.py:671, in KiteConnect._request(self, route, method, parameters)
669 # native Kite errors
670 exp = getattr(ex, data["error_type"], ex.GeneralException)
--> 671 raise exp(data["message"], code=r.status_code)
673 return data["data"]
674 elif "csv" in r.headers["content-type"]:
InputException: interval exceeds max limit: 60 days
To fetch minute-level historical data for a period longer than the 60-day limit imposed by the API, you need to divide your query into multiple smaller queries, each fetching data for up to 60 days. Then, you can stitch these smaller data sets together to create a continuous time series.
Python Script to Fetch and Stitch Minute-Level Historical Data
def fetch_historical_data(kite, token, start_date, end_date, interval):
# Initialize an empty DataFrame to store the historical data
full_data = pd.DataFrame()
# Split the date range into 60-day intervals
while start_date < end_date:
# Calculate the end date of the 60-day interval
interval_end = start_date + datetime.timedelta(days=60)
# Ensure the interval end does not exceed the overall end date
if interval_end > end_date:
interval_end = end_date
# Fetch historical data for the 60-day interval
data = kite.historical_data(token, start_date, interval_end, interval)
data_df = pd.DataFrame(data)
# Append the fetched data to the full dataset
full_data = full_data.append(data_df, ignore_index=True)
# Update the start date for the next interval
start_date = interval_end
return full_data
historical_data = fetch_historical_data(kite, token, starting_date, ending_date, "minute")
historical_data
It took an abnormal amuont of massive time and then it gave output of results with each data of all 454068 trading minutes till date in that range.
date open high low close volume
0 2019-01-01 09:15:00+05:30 1062.55 1063.90 1060.25 1061.65 53427
1 2019-01-01 09:16:00+05:30 1062.25 1063.85 1061.65 1063.40 18209
2 2019-01-01 09:17:00+05:30 1063.40 1063.90 1063.25 1063.40 42785
3 2019-01-01 09:18:00+05:30 1063.40 1063.50 1062.30 1062.60 40647
4 2019-01-01 09:19:00+05:30 1062.30 1062.30 1059.50 1059.50 44316
... ... ... ... ... ... ...
454063 2023-11-24 15:25:00+05:30 2395.95 2396.75 2395.00 2395.65 18228
454064 2023-11-24 15:26:00+05:30 2395.55 2396.00 2394.85 2394.85 12166
454065 2023-11-24 15:27:00+05:30 2395.00 2395.30 2394.50 2394.80 17623
454066 2023-11-24 15:28:00+05:30 2394.55 2394.80 2394.00 2394.00 19879
454067 2023-11-24 15:29:00+05:30 2394.40 2394.90 2394.00 2394.25 9111
454068 rows × 6 columns