I collect data every 30 seconds and store it in a DB. Suppose I want to do analysis for 30m,1H,4H and 1 day time frames, what's the best approach to have access to that data for each TF?

  • 2
    $\begingroup$ Welcome to Quantitative Finance! Your question is very difficult or impossible to answer, since it is much too broad. You need to specify many things: What do you mean by "best approach", what kind of DB and what analysis and so on. If you are asking for specific data structures it could even be off-topic and better posted (after sepcification) to Stackoverflow. $\endgroup$ – g g Oct 7 '17 at 7:26
  • $\begingroup$ Sorry! Maybe my phrase is not accurate enough. I mostly wanted to know how to choose timestamps based on the selected TF and IMHO this may dictate how to store tick data. So Now It seems these are 2 distinct question. $\endgroup$ – Behzad Sedighzadeh Oct 7 '17 at 12:11

Create a separate DB for each time horizon you need in your research.

  • $\begingroup$ A database of End of Day prices normally includes items such as dividends, splits, daily high and low, total day's volume, which are not usually part of a tickdata database. Hence it is best handled as a separate database. $\endgroup$ – Alex C Oct 8 '17 at 3:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.