@RP3436
You can in fact skip celery too. Its not needed. if you are subscribing to just 4-5 instruments, all you need is python, the kiteconnect library, redis and pandas. The efficiencies are all in the way you code. Unfortunately writing code is …
@sauravkedia : sorry for the late response to your reply on 27th feb.
First of all I do not understand why it takes you 2 seconds to load data into pandas.
Zerodha websocket library gives you a list of dictionaries. Each dictionary in the list is a…
@phantomdrake It would help if you write to log file within your app which will help you and others view what is happening with the code internally and at what stage it fails or throws exceptions. The exceptions can be logged too.
This is a sample …
Check for syntax error if you are migrating from kite 2.0 to kite 3.0
def on_ticks( ws , ticks ) : # KIte 3.0
vs
def on_tick( tick , ws ) : # Kite 2.0
And the callback assignment also changes
kws.on_ticks = on_ticks # Kite 3.0
instead of
kws.on…
@pradeepsajjan need to understand a few things about you so that folks can understand and help you better.
A bit about you and where you stand in terms of education, market experience, coding experience etc will help us evaluate your level and help …
This is a programming logic issue.
Are you placing a bracket order or or are you entering a specific stop loss order once your entry is executed? The logic would vary depending on the answer.
You could put a tag on all related orders and then write…
@archulysses I havent tried postgres. In fact I haven't used postgres for years now. However I did read up on it and it seems it supports bson and the performance is pretty good. However whether you want to use postgres or mongo would depend on your…
@vivek What does this error mean
on_close(): ERROR : 1006, connection was closed uncleanly (I dropped the WebSocket TCP connection: not enough arguments for format string)
This has been resolved. Minor issues and RTFM multiple times helped solved the issue.
It was
kws.on_ticks = on_ticks # in kiteconnect 3.0
instead of
kws.on_tick = on_tick # in kiteconnect2.0
@Ashok121: To be safe I always log in at 8:31 or later. Once I started doing that, I never had issues with invalidated access tokens.
That also means that if one wants to do any api related work, one really cant do it for the one hour window between…
Can you give a screenshot of the error? or what is the command you used?
Permission denied when installing any package typically means you are logged in as a regular user and you need to elevate user privileges by appending 'sudo' to whatever comman…
@sujith you mean migrating kite to kite3? :-)
So in simpler terms; until the migration happens although we can use kite3 for all purposes, in order to get the request token we will need to log into kite.zerodha.com. is that understanding correct…
I have tried both. Redis is obviously faster to insert and retrieve data since its in memory.
But what many people do not speak about is that redis converts the data to a utf-8 encoded str. When you read data back in from redis, you need to convert …
This is a late response. But its a bad idea to use historical api in a loop to get a bunch of instrument data. You will use up your 3 calls per second for just getting the data and will not be able to do anything else like get status / place orders …
Instead of redis, you can use mongodb since it supports native object storage and you do not need to convert pickled / str data in redis to python object or what ever programming language you use.
To speed up operations you can configure mongodb to…
NOPE!!!.
Using https://kite.trade/connect/login?v=3&api_key=xxx still redirects me to kite.zerodha.com login although I am already logged in to kite3.zerodha.com.
OK!!
I think I got it. I was trawling through the other posts and ended up with one which had a link to https://kite.trade/docs/connect/v3/user/.
So if I am not mistaken I need to call https://kite.trade/connect/login?v=3&api_key=xxx instead of…
The best way to do this is write your ticks to an in memory database like redis and then use something like python pandas to compute the candles from the tick data.
This involves some learning curve but is superfast though you will be limited by m…
Got it working. For those who would review this thread later, I present my updated code.
order_id = kite.order_place(
exchange="NSE",
tradingsymbol=InstName,
transaction_type="BUY",
…
"For example, we're able to serve a fairly large order book (~100 orders) over HTTP in under 50ms."
Does that mean that there could be an option to get more orderbook (market depth) data?
