yinhm / datafeed Goto Github PK

Hi yinhm

I tried to add the yahoo provider right within the datafeed server loop of my application:

http://pastebin.com/XsBSzZsC

Every time it executes the loop and therefore the whole script terminates right after the yahoo provider has done the job. I can only assume this way to do it is not applicable to the tornado io loop of the datafeed server. Do you have any suggestions? I don't want to place my archive_day function in another script and upload it with client.put_day() anymore.

• fan-out support(?)
• pubsub tick message
• zeromq(?) as a better socks, multiplexing, server push
• websockets(?): fits all morden browsers

Hi @yinhm

I ran across datafeed, and I was wondering if it could be integrated with our project: https://github.com/quantopian/zipline

what do you think?

thanks,
fawce

补好数据之后，发现画出来的图不太对，后来认真看了一下源码，才发现...

兄弟你的这个项目很接地气，找了很久了，但是我水平有限，不太能看得明白怎么进行二次开发。你能弄个文档就好了。
比如，我如何接入网际风接口？如何接收推送来的数据？如何把这些数据存盘（收盘作业？）？
如何取数据进行公式计算？

我试了一下首页的例子：

Client

from datafeed.client import Client
c = Client()
c.get_report("SH000001")

结果是：Exception: Symbol SH000001 not exists.
哪里出错？

• renaming report to tick
• client:
  • get_report -> get_tick
  • get_reports -> get_tickss
  • put_reports -> put_ticks
  • method deprecation, will be remove in the next release.
• all internal naming should change to tick, eg: tickstore

你好，请问有继续开发股票数据的获取，有一些问题想请教你！

• Tick database should be fast enough, use rocksdb until there is another better choice
• Tick store internal, bulk store or just linked list(key)
  • bulk store: deterministic, flush periodic, write-ahead log(?), pick a bulk size fits rocksdb
  • linked list: more simpler logic

python datafeed/server.py

I am working on the integration of a live data feed with zipline. When implementing a new datasource, I found that DataSource is not exported from zipline,sources. Is there some reason for this?
Can I implement a new datasource without subclassing from DataSource? Thanks!