Delta Buy/Sell Volume

[gomifromparis]

if the file size is an issue and you don't use volume filtering, you can use the converter to activate tick compression. that way the file will be smaller and will be parsed faster.

another thing to remember is that the whole file is always read until it finds the first bar timestamp. so if you use a 1 year file only for five days chart, you still read the whole file. so it's a good idea to keep files short. I reset file on contract rollover, and I don't have any perf problems, data shows up in a matter of seconds.

[cwineria]

Hello,

I have a different CD reading on 2 different timeframe charts. As of typing this 1000v reads +31,100 and my 60 minute reads -90,000. If I am not understanding something please help, I do not use historical data and start my feed before the open each day.

Thanks

[paulg]

Hello,

I have a different CD reading on 2 different timeframe charts. As of typing this 1000v reads +31,100 and my 60 minute reads -90,000. If I am not understanding something please help, I do not use historical data and start my feed before the open each day.

Thanks

Could you please post your screens with the examples?

[TheRightSide]

Just thought I will bring this to other CD traders attention out there. CME as made some changes on how transactions are being reported. I believe this would affect CD calculations that use filters based on transaction size.

Emini-watch as written a post on this topic (http://emini-watch.com/cme-says-bett...-think-so/516/)

I'm interested to know any thoughts on how you guys view this change.

[cwineria]

The top chart is a 1000v and the bottom is 60 min. This was taken around 930 est on 10/09/2009

[gomifromparis]

could you show full screen please ?

[cwineria]

could you show full screen please ?

Sure, here you go and thank you.

[gomifromparis]

your time chart must go way sooner than your other chart, so it's picking up some existing data in your file, and calculating a -20000 offset. If you set your both charts on the same starting date you'll have the same results

[gomifromparis]

Here's GomCD 4.1

Fixes/enhancements

* Doesn't plot the 0 values and "big bars" on fast charts any more
* Corrected the crash of indicator when no Gom file existed and you didn't set Write Data to true
* Added a "Forced HiLo bar size" so you can define the bar thickness in Force HiLo mode in non cumulative charts.

[gomifromparis]

Another minor plotting correction that happened sometimes, plotting a strange bar on zero.

[moneybags]

Another minor plotting correction that happened sometimes, plotting a strange bar on zero.

Is there anyway we can get millisecond timestamps on ticks? If its a simple task you could just guide me where to make the change and i can do it. This would help in rebuilding tick based indicators that calculate rate of change, etc.. I've also posted on the RecorderIndicator thread more appropriately.

thanks in advance

[gomifromparis]

Sorry, this isn't possible , because NT has 1 second granularity in timestamps.

[moneybags]

Sorry, this isn't possible , because NT has 1 second granularity in timestamps.

Agreed, the problem lies with NT. However, currently I record the time stamp using DateTime.Now as the ticks are received in real time. I was hoping to use the recorderindicator to store the data. Using some algo you can offset machine time with tick time.

[T2020]

Another minor plotting correction that happened sometimes, plotting a strange bar on zero.

Thanks for the fixes . I'm wondering if there are just some instruments that this just isn't going to work well on . Note before and after pics of the QM .
Second Pic is after F5 refresh . ??

[gomifromparis]

dunno, I double checked NQ and ES and everything is OK.... I don't have access to QM.