Context

Starting something in FinTech is the most difficult thing. You have no open data. These days I'm trying to do some algo-trading. Maybe not in true sense, because it's not high frequency scalping. But anyway that's that.

What?

The data gives almost-Realtime data for half of the Nifty 50 stocks for last week of May and first 2 Weeks of July.

---

Now here is the obvious question. The dataset does not have timestamp. That's because it is collected via Web-Socket streaming as it happens. Sometimes once in a couple of seconds, sometimes 10-15 times in the same span. So there is no point to timestamp IMHO. Anyway it'll be client-side timestamp, so not a true timestamp.

Description

• tick_data.csv contains only the price-volume data.
• volume: total volumes traded for the day
• last_price: denotes the quote price for latest trade
• List item instrument_list.csv contains description of the underlying instrument.

P.S:

**All the data points are not tick-by-tick update. Rather it is mostly an update after 600 ms, provided a trade happened **