Skip to content

Client handling module

The most important part of the algo trading project is the data. For this I am using third party API to donwload the data. Moreover, these APIs can also provide the interfact to place the order on the market. But right now, I will only focus on data downloading part.

There are few data providers like yfinance which are free. Also, there are other clients or say brokers which provides the API for placing orders which also provides historical data, like 5paisa, zeroda, dhan etc. These clients provides the data to their customers free of cost (as far as I know). For this you need to have a Demat account with them. Anyways, I have a demat account on 5paisa and thus I have developed a client using py5paisa api.


flowchart LR

cl[Client] --data request-->
odb[(Client API)] --OHLCV data --> cl
s1[Stock] --as input--> cl

ldb[(local <br>database)]

s1--writes-->ldb
cl--data-->s1
Different clients may have different connection methods. Here, I will show you to how I connect with py5paisa. I keep the API-secrets in yaml config files. I read that and pass it to the Client object. Here is how the config looks like.

account_holder_name: "asdfasdfasdv"
cred:
  APP_NAME: "asdfasasdfadfas"
  APP_SOURCE: "asdfasdfsadfa"
  USER_ID: "asdfasdasdfafasdf"
  PASSWORD: "asdfasdfaasfassdf"
  USER_KEY: "asdfasdghfgfasdf"
  ENCRYPTION_KEY: "asdasfgfasdfacvdsf"
client_code: 'asdfaasfgasdcasc'
mpin: '496165216546'

Loggin-in to the client

Even with the loaded creds, we still need to pass totp. You can get it from a linked authenticator app.

client = fm.clients.client_5paisa.Client5paisa(totp=input("give totp from the autheticator app: "), **creds)

Downloading Data

symbol = "IDEA"
start_date, end_date = "2018-01-01", "2026-01-14"
interval = fm.constants.INTERVAL.one_day

s1 = fm.Stock(symbol)
data = client.download_historical_data(s1,start=start_date, end=end_date, interval=interval)
data.info()

<class 'pandas.core.frame.DataFrame'>
DatetimeIndex: 1992 entries, 2018-01-01 00:00:00 to 2026-01-14 09:15:00
Data columns (total 5 columns):
 #   Column  Non-Null Count  Dtype  
---  ------  --------------  -----  
 0   Open    1992 non-null   float64
 1   High    1992 non-null   float64
 2   Low     1992 non-null   float64
 3   Close   1992 non-null   float64
 4   Volume  1992 non-null   int64  
dtypes: float64(4), int64(1)
memory usage: 93.4 KB

Note that this is one_day interval data.