Cohort Share Discussion Threads

[bigjacko] 21/4/2004 21:56

LOL, Hooya... and there was me thinking "Now he's got that sorted, I wonder whose backside we have to kiss in order to get VOLUME data for the UKX?"

Perhaps I'll wait until after you've asked yours. It is your thread, after all :o)

[hooya] 21/4/2004 19:06

David I have some questions for you on the skew bands tomorrow...so better do your homework :-)

[bigjacko] 21/4/2004 19:02

Yes, I meant really that rather than just slipping the option out there on the quiet, it might be wise for ADVFN to announce the option, and the reasons behind the change, really. Some people get rather committed to things 'the way they are' simply because, well, that's the way they are! LOL.

Without education, they may not know what they are missing, is all I'm getting at. I'd hate for all your and David's hard work and efforts to go completely unnoticed amongst the zillions of other 'options' available here. This one is important, imho.

[hooya] 21/4/2004 18:39

" putting it in as an option is a very sensible idea, I agree. We'll have to work on putting together some information, so people know WHY one method might be better than the other, though."

I still cant see why anyone would want a candle chart not to dislpay the close or the open as in at least one of them?

But agree pay rise but after its been live for a bit:-)

[fross] 21/4/2004 13:47

oops, forgot to add something else to that post.

i'm going to add this in as an *option* in the chart, so you can choose whether to watch strict market hours, or to include post/pre-market activity. so nobody loses out and the behaviour doesn't suddenly change. i'm still playing about with representing the last bar differently but it might not be necessary.

this won't go live quite yet but it's getting there, just needs to go through some very rigorous testing as it's a potentially big change. next week on beta, i hope.

DavidW

[hooya] 21/4/2004 13:44

"the last candle matches the yesterday's close line. the open of the following day does not (necessarily) start from the YC "


WOO HOO!!!!!!!!!!!!!!!!!!!:-)

[hooya] 20/4/2004 17:23

Oi all three of you can go and have arguements on the FBB but not on this thread please

[hooya] 20/4/2004 14:37

big charts are pretty close too

[neo anderson] 20/4/2004 14:37

LOLOL!!!!!

( real one ) (as endorsed by Schroders)

[fross] 20/4/2004 14:35

well we can swap the colours round if it'll make that big of a difference! ;)

[hooya] 20/4/2004 14:28

I'll be honest I thought there would/might be a bigger difference but look at this

[neo anderson] 20/4/2004 14:21

I agree David.

Mad. Where's your reference material from which you've considered your query?

Thanks.

[fross] 20/4/2004 14:14

mad4IT,

i'm happy to discuss formulae and algorithms we use, and we can compare what we currently have implemented with other reference designs for analysis tools. and yes, if there is a significant difference i have no problem with us having two MACD tools (for instance) with both implementations running, as there's twice as much use in twice the number of tools! the help pages give some basic information on how the studies work, but i can provide more details for discussion - i suggest you point us at some reference material for MACD or RSI and i can then reply with how things may differ.

DavidW

[hooya] 20/4/2004 14:13

should be easy to check

RSI= 100-(100/(1+RS)) or =100 * (RS/(1+RS))

where RS = total higher closes/total lower closes in the perriod selected

So is that the formula ADVFN use?

[neo anderson] 20/4/2004 14:03

Mad.

The formulae for RSI and MACD are constant wherever you read them (or they should be). You've looked at two sites that may use the same provider for their source data. ADVFN aren't wrong as I'm sure they have used the correct formula. If there is a problem it will be in the disparity of the data providers and IMO you're looking like a bit of a plum for pulling them up on it when anyone with half a clue would know that the constituent factors of a formula are constant whereas different results will be seen if the source data is different.

Mad mad mad. Where did you go to school?

Regards.

Obvious Anderson

[mad4it] 20/4/2004 12:36

Cheers David.

I've compared the AFN RSI & MACD data to 'Big Charts', Updata, etc and while they concur, AFN differs. I'm assuming it's down to the initial calculation( alogorithms?) that were inputed onto the system a couple of years ago ?

Can I just clarify this:

" don't see a problem with using a "standard" RSI/MACD and keeping our current ones as differently named versions, or even vice versa. i'm mostly interested in implementing it according to the best referenced standard. "

I take it you're suggesting two versions of the RSI & MACD on AFN ?

If so, I think that's an excellent idea, because I have noticed that the 'incorrect' calculation on AFN can occasionally be more useful than the 'standard' calculation elsewhere. This is probably explained by MMs using the standard data to fake out traders and the AFN data can be a useful comparison in studying historical price movements.