Creating a trading system from scratch

[vinit_7801]

If you can place orders from mobile, then you can use my website.
Message me for details

in site can you add just one more fic...to choose the number of stocks...that would be great

 

[UberMachine]

in site can you add just one more fic...to choose the number of stocks...that would be great

Nice suggestion. Already planned. Would come up in the upcoming version

 

[VJAY]

Zerodha provides an offline basket order execution so that I can place an order via browser. Since browsers can operate anywhere, you can place orders from any platform.

I currently use google compute for auto placing orders at 9:15:00 even when not available


I run only on Linux
Just use any broker who provides you an interface with browser.
Zerodha provides them (Upstox should also provide it but got to check)
Else you need to use an API but API's are costly for retail traders, in my opinion. If you have an extremely well tested scalping strategy, you can use it.

You could run a scheduled task by converting the notebook into a python file (so that you get the output).
For now, just export the notebook as python script to enable this and try it out

Dear UB can you please explain bold part?

 

[UberMachine]

Dear UB can you please explain bold part?

Convert your existing notebook to an Executable script.

Now you could use task scheduler to run this task at a specified time daily.
One important thing is we got to change the filename daily before the script runs.
You can try it now so that you get used to it before I automate even that part.

 

[Shivam_]

This might reduce some effort of downloading csv and changing dates
Edit-Forgot Holidays now must work fine
still needs to consider if two holiday are together except SAT,SUN maybe creating local file having holiday dates was less messy

import os.path
import requests,zipfile,io
import datetime
today = datetime.date.today()
yesterday = today - datetime.timedelta(days = 10)
PREVDAY=yesterday.strftime("%a")

if PREVDAY=='Sun':
    yesterday -= datetime.timedelta(days = 2)

DATE=yesterday.strftime("%d%b%Y").upper()
YEAR=yesterday.strftime("%Y")
MON=yesterday.strftime('%b').upper()

LINK= f"https://www.nseindia.com/content/historical/EQUITIES/{YEAR}/{MON}/cm{DATE}bhav.csv.zip"

print(f'Downloading -{DATE} csv')

r = requests.get(LINK)

if len(r.content)<300:
    print('Yesterday was holiday')
    yesterday -= datetime.timedelta(days = 1)
    DATE=yesterday.strftime("%d%b%Y").upper()
    YEAR=yesterday.strftime("%Y")
    MON=yesterday.strftime('%b').upper()
    LINK=f"https://www.nseindia.com/content/historical/EQUITIES/{YEAR}/{MON}/cm{DATE}bhav.csv.zip"
    r = requests.get(LINK)
    z = zipfile.ZipFile(io.BytesIO(r.content))
    z.extractall(path='C:\\tradingsystem\\data')
    print (f"Check if ORDERFILE {DATE} date is correct, File downloaded")

else:
    z = zipfile.ZipFile(io.BytesIO(r.content))
    z.extractall(path='C:\\tradingsystem\\data')
    print ("File downloaded")

ORDER_FILENAME = f'data/cm{DATE}bhav.csv'

 

[vinit_7801]

This might reduce some effort of downloading csv and changing dates

import os.path
import requests,zipfile,io
import datetime
today = datetime.date.today()
yesterday = today - datetime.timedelta(days = 1)
DATE=yesterday.strftime("%d%b%Y").upper()
YEAR=yesterday.strftime("%Y")
MON=yesterday.strftime('%b').upper()
LINK= f"https://www.nseindia.com/content/historical/EQUITIES/{YEAR}/{MON}/cm{DATE}bhav.csv.zip"
print(f'Downloading -{DATE} csv')
if os.path.isfile(f"data/cm{DATE}bhav.csv")
    print ("File exists and is readable")
else:
    r = requests.get(LINK)
    z = zipfile.ZipFile(io.BytesIO(r.content))
    z.extractall(path='C:\\tradingsystem\\data')
    print ("File downloaded")

ORDER_FILENAME = f'data/cm{DATE}bhav.csv'

why download yesterday's bhavcopy any specific reason??
And should i change the systemnew orderfile as well..will it work?

 

[Shivam_]

why download yesterday's bhavcopy any specific reason??
And should i change the systemnew orderfile as well..will it work?

hain
we are using 10oct csv result on 11 oct to trade if i'm correct
so if you use this on 12 it will download 11oct csv
orderfile will have yesterday date in filename if you use this automatically

 

[vinit_7801]

hain
we are using 10oct csv result on 11 oct to trade if i'm correct
so if you use this on 12 it will download 11oct csv
orderfile will have yesterday date in filename if you use this automatically

Got it...it is schedule for tommorrow 9:12 right...
tried it on python works like a charm...hope the scheduler works tomorrow

 

[Shivam_]

Got it...it is schedule for tommorrow 9:12 right...
tried it on python works like a charm...hope the scheduler works tomorrow

Schedule is to be done on windows scheduler not in program code
i got confused if you got confused

 

[vinit_7801]

Schedule is to be done on windows scheduler not in program code

Ya i m using windows 10 n have schedule it for tommorrow 9.12...Hope it works fingerscrossed...