Chat Session Logs

4:34 pm:<Duke> Hey Chad...
4:34 pm:<LS_Chad> Hey Duke
4:36 pm:<Steve> thx duke got the definition!
4:37 pm:<LS_Chad> Anyone ever tried to code the Hammer or Inverse Hammer in a signal?
4:37 pm:<Steve> Well without a DOUBT I am the ranking newbie at the moment!!!
4:38 pm:<Steve> Chad your charts still open perfectly with the pane sizes exactly as you left them the data before?
4:39 pm:<LS_Chad> yea...just opened again...
4:39 pm:<Steve> If so try leaving some minimized in your layout and then open one of those the next day from minimized in your saved layout and I am sure the bottom pane will have enlarged.
4:40 pm:<Steve> and ?
4:40 pm:<LS_Chad> no problem...same as original
4:40 pm:<Steve> try what I suggested above then
4:41 pm:<LS_Chad> ok
4:41 pm:<F_Trader> Hi, just back
4:42 pm:<LS_Chad> As far as the Hammer...it looks like the lower shadow should be at least twice as long as the body...and the upper shadow should be very small or non-existent
4:43 pm:<F_Trader> hammer something like that indeed,
4:43 pm:<LS_Chad> to get body...
4:43 pm:<LS_Chad> ABS(CL - OP)
4:44 pm:<LS_Chad> lower candle...
4:44 pm:<LS_Chad> SMIN(OP, CL) - LO
4:44 pm:<LS_Chad> sorry...lower shadow
4:44 pm:<LS_Chad> upper shadow...
4:44 pm:<LS_Chad> HI - SMAX(OP, CL)
4:44 pm:<F_Trader> still ok to me
4:45 pm:<LS_Chad> and I would assume the range (HI - LO) should be some minimum (based on maybe the average of previous 20 bars...say at least half)..
4:46 pm:<LS_Chad> (HI - LO) > (0.50 * AVG(HI - LO, 20))
4:46 pm:<LS_Chad> lower shadow at least twice body...
4:46 pm:<F_Trader> logical
4:47 pm:<LS_Chad> (SMIN(OP, CL) - LO) >= (2 * ABS(CL - OP))
4:47 pm:<LS_Chad> upper shadow minimal on non-existent (we'll say has to be less than 6% of body I guess)
4:48 pm:<LS_Chad> (HI - SMAX(OP, CL)) <= (0.06 * (HI - LO))
4:48 pm:<LS_Chad> all together...
4:48 pm:<F_Trader> you got all the rules
4:48 pm:<LS_Chad> ((HI - LO) > (0.50 * AVG(HI - LO, 20))) AND ((SMIN(OP, CL) - LO) >= (2 * ABS(CL - OP))) AND ((HI - SMAX(OP, CL)) <= (0.06 * (HI - LO)))
4:49 pm:<LS_Chad> since I've got no new users here...just talking to myself ;)
4:49 pm:<F_Trader> Chad,
4:49 pm:<F_Trader> I saw Mike already wrote
4:50 pm:<F_Trader> about removing color purple and * from hard coded text to chat
4:50 pm:<F_Trader> so that text would look like usual post
4:51 pm:<LS_Chad> yea..can do...
4:51 pm:<LS_Chad> what's the * though...don't recall that...
4:51 pm:**401** LS_Chad test :No such nick
4:51 pm:<F_Trader> i think this come from using /msg
4:52 pm:* F_trader01 TESTING
4:52 pm:<F_Trader> the green text is from me
4:52 pm:<LS_Chad> right..ok....I kinda like that because it stands out from other messages....why do you want that removed?
4:52 pm:* F_trader01 TESTING
4:53 pm:<F_Trader> is what u get by default
4:53 pm:<LS_Chad> right..I'm in IRT so I don't see any colors
4:54 pm:<F_Trader> ok but u see the *
4:54 pm:<F_Trader> in front of the nickname
4:55 pm:<F_Trader> instead of < >
4:57 pm:<LS_Chad> yes
4:57 pm:<F_Trader> ok
4:58 pm:<LS_Chad> I'm just asking why you prefer no *, as opposed to *
4:58 pm:<F_Trader> i was just posting thru text to chat in another forum, and was asked to post as everyone else, without purple and *
4:59 pm:<F_Trader> indeed, with all my post in purple, this was quite embarrassing
5 pm:<Duke> he is afraid folks will start calling him violet!! :-)
5 pm:<F_Trader> and being with a * in front of them, while everybody else was posting by typing
5 pm:<F_Trader> lol Duke
5 pm:<F_Trader> Mike told me this was hard coded inside IRT
5:01 pm:<LS_Chad> yea...I understand....ok
5:01 pm:<LS_Chad> might want to remind me tomorrow to look into that..
5:01 pm:<LS_Chad> this signal is very processor intensive for some reason
5:01 pm:<LS_Chad> ((HI - LO) > (0.50 * AVG(HI - LO, 20))) AND ((SMIN(OP, CL) - LO) >= (2 * ABS(CL - OP))) AND ((HI - SMAX(OP, CL)) <= (0.06 * (HI - LO)))
5:01 pm:<F_Trader> no problem
5:02 pm:<F_Trader> SMIN probably, a CL always change
5:02 pm:<F_Trader> as
5:03 pm:<F_Trader> i would replace everything by .1 by the way, to make sure this is a confirmed hammer
5:03 pm:<LS_Chad> no...it's AVG(HI - LO, 20)
5:03 pm:<F_Trader> this would speed calc speed
5:03 pm:<F_Trader> AVG being a STAT
5:03 pm:<F_Trader> possible indeed
5:03 pm:<F_Trader> did u tried with HLR ?
5:03 pm:<F_Trader> could speed up may be things
5:04 pm:<LS_Chad> Yes...HLR (or using STAT) works much better...
5:04 pm:<F_Trader> or just try thru a MA like 20*MA(HLR)
5:05 pm:<LS_Chad> probably ought to add one more condition to this...require the hammer head (the body) to be some minimal size...
5:05 pm:<LS_Chad> like 10 % of the range
5:06 pm:<F_Trader> yep, if there is no body, ie CL=OP, then this specific hammer has another name
5:06 pm:<F_Trader> not sure which one
5:06 pm:<LS_Chad> ((HI - LO) > (0.50 * AVG(HLR, 20))) AND ((SMIN(OP, CL) - LO) >= (2 * ABS(CL - OP))) AND ((HI - SMAX(OP, CL)) <= (0.06 * (HI - LO))) AND (ABS(CL - OP) > (0.10 * (HI - LO)))
5:07 pm:<LS_Chad> doji I think...something like that
5:07 pm:<F_Trader> gravestone doji
5:07 pm:<LS_Chad> that CI is still taking about 266ms...not real happy with that
5:07 pm:<F_Trader> you know Chad, CI time is really depending on how much data u have in db
5:08 pm:<F_Trader> and another tip : on my side , i never never use AND
5:08 pm:<LS_Chad> well...I think that is true...for some CI's...
5:08 pm:<LS_Chad> it depends on which tokens you're using...
5:08 pm:<F_Trader> i always use a *
5:08 pm:<LS_Chad> like MPD...or SSTAT...
5:08 pm:<LS_Chad> why use *?
5:09 pm:<F_Trader> ((HI - LO) > (0.50 * AVG(HLR, 20))) * ((SMIN(OP, CL) - LO) >= (2 * ABS(CL - OP))) * ((HI - SMAX(OP, CL)) <= (0.06 * (HI - LO))) * (ABS(CL - OP) > (0.10 * (HI - LO))) >0
5:09 pm:<F_Trader> much quicker
5:09 pm:<F_Trader> if you have AND or OR
5:09 pm:<F_Trader> it slow down speed calc
5:09 pm:<F_Trader> in all my signals, i always have * or +
5:10 pm:<LS_Chad> interesting
5:10 pm:<F_Trader> A true or B true will de bore me (A+B)>0
5:10 pm:<F_Trader> with A being a signal
5:10 pm:<F_Trader> the ultimate example of that is
5:11 pm:<F_Trader> (A or B) true AND (C or D) true
5:11 pm:<F_Trader> if your start mix AND and OR
5:11 pm:<F_Trader> it is a nightmare for calc time
5:11 pm:<LS_Chad> ever mentioned this to Bill?
5:12 pm:<F_Trader> but ((A+B)>0 ) + ((C+D)>0) > 0
5:12 pm:<F_Trader> works as quick as any other CI
5:12 pm:<F_Trader> not recently, i think i mentioned that a yr ago or so
5:13 pm:<F_Trader> but u may just email hin this piece of log..
5:14 pm:<F_Trader> this is by the way a great advantage of using mirc : automating logging of all your #linnsoft chat...
5:14 pm:<F_Trader> oops sorry
5:14 pm:<F_Trader> above syntax incorrect
5:14 pm:<F_Trader> this should be ((A+B)>0 ) * ((C+D)>0) > 0
5:15 pm:<F_Trader> for (A or B) true AND (C or D) true
5:16 pm:<F_Trader> do u get lower calc time with a *
5:16 pm:<LS_Chad> I changed my Hammer and it didn't seem to make much difference..
5:16 pm:<F_Trader> Indeed, the gain in speed is higher when a + replaces an OR statement
5:17 pm:<F_Trader> impact far less if u only have AND
5:17 pm:<F_Trader> but all my comments were true before Bill made this change to calc spped
5:18 pm:<F_Trader> overall calc speed has improved significantly since few months
5:18 pm:<F_Trader> Chad : how do you check calc speed in this case ????
5:18 pm:<F_Trader> inside a chart ?
5:19 pm:<F_Trader> if this is the case, u need to reboot IRT
5:19 pm:<F_Trader> before getting a next correct calc time
5:20 pm:<F_Trader> otherwise, if formula name unchanged, IRT will take the calc time of past formula inside new calc speed value
5:20 pm:<F_Trader> found this little surprise few weeks ago
5:20 pm:<LS_Chad> right..you can just close and reopen chart...
5:21 pm:<F_Trader> not enough !!!
5:21 pm:<LS_Chad> or remove and re-add indicator....
5:21 pm:<F_Trader> may be remove and re-add, but not sure
5:21 pm:<F_Trader> remember my email with the 2850 ms post time
5:22 pm:<LS_Chad> I have recommended to Bill that we change the value shown to just reflect the time of "last" calc...instead of the average...which can be skewed
5:22 pm:<F_Trader> fully agree
5:22 pm:<F_Trader> will be more representative
5:22 pm:<F_Trader> or may be take last 5 calc time
5:22 pm:<LS_Chad> yes...I don't think that average calc (or total calc time for that matter) is really of much value
5:22 pm:<LS_Chad> just need last calc time
5:23 pm:<F_Trader> good idea
5:23 pm:<F_Trader> Chad, got two little display issue for you
5:24 pm:<F_Trader> the first one is
5:24 pm:<F_Trader> select a chart and push the delete button
5:24 pm:<F_Trader> this make the list of element appear
5:24 pm:<F_Trader> i always miss half the name of the elements
5:25 pm:<F_Trader> if i make the window wider
5:25 pm:<F_Trader> this won t help
5:26 pm:<LS_Chad> some indicators with really long names?
5:26 pm:<F_Trader> posting one example
5:26 pm:<F_Trader> one sec
5:27 pm:<F_Trader> one sec
5:28 pm:<F_Trader> www.ef1970.easynet.be/example.doc
5:28 pm:<F_Trader> not here yet
5:29 pm:<F_Trader> yep
5:30 pm:<F_Trader> main problem are color marker
5:30 pm:<F_Trader> as i got 800 CI, i need to classify them using same initial letters
5:31 pm:<F_Trader> like CI TRADE_STATE_LONG_Entry_point CI_TRADE_STATE_LONG_exit_point etc
5:32 pm:<F_Trader> but then can't find who is who
5:32 pm:<X> Welcome to Linn Software's Investor/RT.
5:33 pm:<F_Trader> I got same problem for the "select graph element" window
5:37 pm:<F_Trader> Chad plunged into some code ?
5:40 pm:<F_Trader> most probably :-)
5:41 pm:<LS_Chad> was just cranking out another email on this Hammer signal....
5:41 pm:<LS_Chad> one sec
5:41 pm:<F_Trader> no problem
5:45 pm:<LS_Chad> I think it shows you like 32 characters there...
5:45 pm:<F_Trader> which is the max for a CI or SIG name
5:46 pm:<F_Trader> however, if you get color marker ES M3 and a long CI you miss some
5:46 pm:<F_Trader> did u get my screenshot ?
5:47 pm:<F_Trader> www.ef1970.easynet.be/example.doc
5:47 pm:<F_Trader> it shows as well another little play i do very often
5:47 pm:<LS_Chad> yea...I see what you're saying...
5:48 pm:<F_Trader> if you could make it 50 or so
5:48 pm:<F_Trader> would fit
5:48 pm:<LS_Chad> so you want the name to come first maybe...
5:48 pm:<LS_Chad> like CI does
5:48 pm:<F_Trader> good idea
5:49 pm:<F_Trader> but i could then miss the instrument :)
5:49 pm:<LS_Chad> COLOR:NAME ....other stuff
5:49 pm:<LS_Chad> yea
5:49 pm:<F_Trader> and have same color marker same CI for different instrument as well
5:49 pm:<LS_Chad> I'll talk to Bill about this...(he's out of town currently)
5:49 pm:<F_Trader> anyhow nothing urgent
5:49 pm:<F_Trader> got same issue with CI window
5:50 pm:<F_Trader> making t wider do not wide element list window
5:50 pm:<F_Trader> it
5:50 pm:<F_Trader> see the example ?
5:50 pm:<LS_Chad> right.....we could make that list on the right widen when the window is widened
5:51 pm:<F_Trader> exactly
5:51 pm:<F_Trader> always move middle section so i may read successively CI and TSIG name
5:52 pm:<F_Trader> not a big deal but would help save some time while checking CI
5:52 pm:<Steve> in the scan window the QP dropdown menu is not wide enough either to view the names of qps completely
5:53 pm:<Steve> if they are long names
5:54 pm:<F_Trader> true indeed Steve
5:54 pm:<Steve> and the token window does not display the whole token name
5:54 pm:<F_Trader> I am to lazy to have long token name :)
5:55 pm:<F_Trader> but got similar problem in picking a signal name inside a signal marker definition window
5:55 pm:<LS_Chad> Good suggestions guys...I'll see if we can get them implemented...
5:56 pm:<F_Trader> cosmetic stuff
5:56 pm:<Steve> I am talking about reading the built in token names in this case...even when the window is expanded many names are only partially readable
5:56 pm:<Steve> because that section does not expand
5:56 pm:<F_Trader> but cosmetic may matter
5:57 pm:<LS_Chad> Steve..you know you can double click on a column heading (token column heading) to have the column fit the tokens within it
5:57 pm:<F_Trader> check my wife "cosmetic's" budget :-=)
5:57 pm:<Steve> for us perennial newbies it is nice to read the descriptive token name
5:58 pm:<Steve> to decide if it might do the trick or not
5:58 pm:<LS_Chad> oh...the description...I see
5:58 pm:<F_Trader> Chad : not in the pickup list
5:59 pm:<Steve> speaking of cosmetics ...the info window in the lower panes displaying data is great but the date and time are unnecessary and take up vertical space in small panes
5:59 pm:<LS_Chad> yea...maybe when you click on it...the description should show up somewhere else...like window title..