A Doji candlestick is a special type of candlestick pattern in the stock market that shows indecision between buyers and sellers. When doji candlestick forms, it indicates that the market opens and closes at the same price. This means there is equality and indecision between buyers and sellers, with no one controlling the market

How a Doji Candle Looks

• Very small or no body (because open ≈ close)
• Long wicks (shadows) may appear on the top or bottom

from kiteconnect import KiteConnect
import pandas as pd
import datetime as dt
import os
import numpy as np

cwd = os.chdir("D:\GenAi\python\python basic\Strategy")

access_token = open("access_token.txt",'r').read()
key_secret = open("api_key.txt",'r').read().split()
kite = KiteConnect(api_key=key_secret[0])
kite.set_access_token(access_token)


instrument_dump = kite.instruments("NSE")
instrument_df = pd.DataFrame(instrument_dump)

def instrumentLookup(instrument_df,symbol):
    try:
        return instrument_df[instrument_df.tradingsymbol==symbol].instrument_token.values[0]
    except:
        return -1

def fetchOHLC(ticker,interval,duration):
    instrument = instrumentLookup(instrument_df,ticker)
    data = pd.DataFrame(kite.historical_data(instrument,dt.date.today()-dt.timedelta(duration), dt.date.today(),interval))
    data.set_index("date",inplace=True)
    return data

def doji(ohlc_df):    
    """returns dataframe with doji candle column"""
    df = ohlc_df.copy()
    avg_candle_size = abs(df["close"] - df["open"]).median()
    df["doji"] = abs(df["close"] - df["open"]) <=  (0.05 * avg_candle_size)
    return df


ohlc = fetchOHLC("TATAMOTORS","5minute",5)
doji_df = doji(ohlc)

Line-by-Line Explanation

1. Importing Libraries

from kiteconnect import KiteConnect
import pandas as pd
import datetime as dt
import os
import numpy as np

• KiteConnect → Used to connect with Zerodha Kite API.
• pandas → Used to create dataframes and handle market data.
• datetime → Used to calculate date ranges for historical data.
• os → Used for file/folder path settings.
• numpy → Used for numeric operations (optional here).

2. Change Current Working Directory

cwd = os.chdir("D:\GenAi\python\python basic\Strategy")

• Changes the working folder to your strategy directory.
• So Python will look for files (access_token.txt, api_key.txt) inside this folder.

3. Generate Trading Session

access_token = open("access_token.txt",'r').read()

Opens the file access_token.txt and reads the access token (used for authenticated API calls).

key_secret = open("api_key.txt",'r').read().split()

Reads API key & secret from api_key.txt and splits into a list.
Example: ["your_api_key", "your_api_secret"]

kite = KiteConnect(api_key=key_secret[0])

Creates a KiteConnect object with your API key.

kite.set_access_token(access_token)

Sets the access token so you can call private API endpoints.

4.Get all NSE instruments

instrument_dump = kite.instruments("NSE")

• Fetches all tradable instruments from NSE exchange.
• Zerodha returns a big list of dictionaries, including instrument names, tokens, etc.

instrument_df = pd.DataFrame(instrument_dump)

• Converts that list into a pandas DataFrame for easy lookup.

5. instrument Lookup()

def instrumentLookup(instrument_df,symbol):
    try:
        return instrument_df[instrument_df.tradingsymbol==symbol].instrument_token.values[0]
    except:
        return -1

• Searches for the trading instrument by symbol (e.g., “TCS”, “INFY”).
• Returns its instrument_token, which is required for historical API.
• If symbol not found → returns -1.

6. fetchOHLC()

def fetchOHLC(ticker,interval,duration):
    instrument = instrumentLookup(instrument_df,ticker)

• Gets the instrument token for the ticker.

 data = pd.DataFrame(kite.historical_data(
        instrument,
        dt.date.today()-dt.timedelta(duration),
        dt.date.today(),
        interval))

• Fetches historical OHLC data using:
• instrument token
• start date (today – duration days)
• end date (today)
• interval (e.g., “5minute”, “day”)

   data.set_index("date", inplace=True)

• Makes the date column the index of the DataFrame.

 return data

• Returns the complete OHLC dataset.

7.doji() — Identify Doji Candlestick

def doji(ohlc_df):    
    df = ohlc_df.copy()

• Creates a copy of the input OHLC DataFrame.

 avg_candle_size = abs(df["close"] - df["open"]).median()

• Calculates median body size of all candles.
• A doji has a very small body.

 df["doji"] = abs(df["close"] - df["open"]) <= (0.05 * avg_candle_size)
    return df

• Marks a candle as doji when body ≤ 5% of median candle size.
• Creates a new column "doji" with True/False.
• Returns DataFrame with doji information.

8.Execute the program and Identify Doji candles

ohlc = fetchOHLC("TATAMOTARS","5minute",5)
doji_df = doji(ohlc)