| StockFetcher Forums · General Discussion · Debug Issue with: "Warning: period length exceeded." | << 1 2 >>Post Follow-up |
| jimhbutton 104 posts msg #66886 - Ignore jimhbutton modified |
9/2/2008 11:12:31 AM Howdy, My filter seems to work without issue till I plug in certain SYMLIST candidates that appear to fit the price & volume criteria of those that passed muster. Can anyone take a look at this filter and tell me why it is fetching hundreds while giving the following glitch debug phrase ("Warning: period length exceeded.") for other SYMLIST-ers such as ECPG or WNR that appear worthy? Fetcher[ /* 1 -- Master Run Prognosticator (Combined -- Previous-Close and Open-to-High-Today) */ /* Note: Accurate Info Once StockFetcher's Initial Market Update Around 10 AM */ set{Hi1, High - Close 1 day ago} set{AvgClRun3, cma(Hi1, 3)} set{AvgClRun5, cma(Hi1, 5)} set{AvgClRun50, cma(Hi1, 50)} set{AvgClRun100, cma(Hi1, 100)} set{AvgClRun275, cma(Hi1, 275)} set{AvgClRun450, cma(Hi1, 450)} set{ClHiLo5, Hi1 5 day Low} set{ClHiHi5, Hi1 5 day High} set{ClRun-TgtClHiLo5, Close 1 day ago + ClHiLo5 } set{ClRun-TgtClHiHi5, Close 1 day ago + ClHiHi5} set{ClRun-Tgt3, Close 1 day ago + AvgClRun3} set{ClRun-Tgt5, Close 1 day ago + AvgClRun5} set{ClRun-Tgt50, Close 1 day ago + AvgClRun50} set{ClRun-Tgt100, Close 1 day ago + AvgClRun100} set{ClRun-Tgt275, Close 1 day ago + AvgClRun275} set{ClRun-Tgt450, Close 1 day ago + AvgClRun450} set{HiOp1, High - Open} set{AvgOpRun3, cma(HiOp1, 3)} set{AvgOpRun5, cma(HiOp1, 5)} set{AvgOpRun50, cma(HiOp1, 50)} set{AvgOpRun100, cma(HiOp1, 100)} set{AvgOpRun275, cma(HiOp1, 275)} set{AvgOpRun450, cma(HiOp1, 450)} set{HiLo5, HiOp1 5 day Low} set{HiHi5, HiOp1 5 day High} set{OpRun-TgtHiLo5, Open + HiLo5} set{OpRun-TgtHiHi5, Open + HiHi5} set{OpRun-Tgt3, Open + AvgOpRun3} set{OpRun-Tgt5, Open + AvgOpRun5} set{OpRun-Tgt50, Open + AvgOpRun50} set{OpRun-Tgt100, Open + AvgOpRun100} set{OpRun-Tgt275, Open + AvgOpRun275} set{OpRun-Tgt450, Open + AvgOpRun450} /* DISPLAY COLUMNS */ add column separator add column separator add column ClRun-Tgt450 {C-R450} add column ClRun-Tgt275 {C-R275} add column ClRun-Tgt100 {C-R100} add column ClRun-Tgt50 {C-R50} add column ClRun-TgtClHiHi5 {[ Hi5} add column ClRun-Tgt5 {C-R5} add column ClRun-TgtClHiLo5 {Lo5 ]} add column ClRun-Tgt3 {C-R3 =} add column separator add column AvgClRun450 {+450} add column AvgClRun275 {+275} add column AvgClRun100 {+100} add column AvgClRun50 {+50} add column AvgClRun5 {+5} add column AvgClRun3 {+3} add column separator add column Close 1 day ago {YCLOSE} add column separator add column separator add column Open add column separator add column AvgOpRun3 {+3} add column AvgOpRun5 {+5} add column AvgOpRun50 {+50} add column AvgOpRun100 {+100} add column AvgOpRun275 {+275} add column AvgOpRun450 {+450} add column separator add column OpRun-Tgt3 {= O-R3} add column OpRun-TgtHiLo5 {[ Lo5} add column OpRun-Tgt5 {O-R5} add column OpRun-TgtHiHi5 {Hi5 ]} add column OpRun-Tgt50 {O-R50} add column OpRun-Tgt100 {O-R100} add column OpRun-Tgt275 {O-R275} add column OpRun-Tgt450 {O-R450} add column separator add column separator /* SELECTION CRITERIA */ Price < 15 and Average Volume (30) > 100000 and draw Linear Regression Indicator(20) and draw Parabolic SAR(0.02,0.2) and draw Parabolic SAR(0.04,0.2) and draw Parabolic SAR(0.06,0.2) and draw MA(50) and draw MA(200) SYMLIST() /* SORT CRITERIA */ SORT COLUMN 29 DESCENDING] |
| chetron 2,817 posts msg #66887 - Ignore chetron modified |
9/2/2008 12:28:08 PM clickable... excessive brackettes |
| maxreturn 745 posts msg #66888 - Ignore maxreturn modified |
9/2/2008 12:29:27 PM Chet to the rescue! |
| chetron 2,817 posts msg #66889 - Ignore chetron modified |
9/2/2008 12:59:20 PM LOL ACTUALLY, MAX, YOUR ORIGINAL POST IS THE CORRECT ANSWER TO HIS PROBLEM. THE ERRORS I FOUND ONLY HAD TO DO WITH CLICKABLITY. THE MAX PERIOD IS 100 FOR EVERYONE. THERE ARE WAYS TO TRICK IT AS TRO HAS POSTED MANY TIMES. SEARCH THE FORUMS. |
| jimhbutton 104 posts msg #66890 - Ignore jimhbutton modified |
9/2/2008 1:50:23 PM Chet, Thank you for taking the time to help me. I appreciate it. Was there an error with a single extra bracket, or did I exceed the overall allowable number of brackets? I'm comparing my original with your amended version and I've yet to pinpoint my mistake. jhb2 P.S. I see it now... The Lo5 and Hi5 brackets... Is the data over 100 days accurate? |
| chetron 2,817 posts msg #66891 - Ignore chetron |
9/2/2008 1:54:05 PM MAYBE MY MATH IS INVALID, BUT THE PRINCIPLE ISN'T... |
| chetron 2,817 posts msg #66892 - Ignore chetron |
9/2/2008 1:57:53 PM SIMPLER..... |
| jimhbutton 104 posts msg #66894 - Ignore jimhbutton modified |
9/2/2008 2:38:16 PM Chet, Tro, Niko, et al: Chetron do you know which one of your revisions gives ACCURATE results beyond the typical 100 days? Running all three revised scans for WNR I received three different sets of answers for the 100+ day columns: C-R450: 9.46 9.65 9.62 C-R275: 9.70 9.96 9.55 O-R275: 9.92 10.15 9.80 O-R450: 9.69 9.91 9.87 In my original filter at the beginning of this thread DIFFERENT results are retrieved for every cma(xxx, yyy) count where the yyy amounts range from 101 through 450 days. After 450 days the results were redundant, leading one (erroneously?) to believe that the amounts up to then were ACCURATE. If in my original filter they are not accurate then just what do my 101 - 450 day results reflect? Has anyone validated the ACCURACY of the 100+ day figures in any 100+ day system of computation, mine included? The reason I ask this is because in the following link posted on July 15, 2003 @ http://forums.stockfetcher.com/sfforums/?q=view&fid=1004&tid=27374&qrid=&isiframe= StockFetcher.com stated: "count(screening phrase , period) "Current Limitations: "You can only include a single screen within the count(...) text. That is, you can't include filters which have been joined with "and" inside of the count(...) text. "THERE IS A LIMIT OF 100 DAYS for accumulating the counts. This restriction is based on performance and HOPEFULLY WE WILL BE ABLE TO INCREASE THAT VALUE." My question, five years later: Was StockFetcher "able to increase that value"? And has anyone verified it since 2003? Before I revise my present filters (with their 101-day to 450-day feature) with one proven to be ACCURATE I would like confirmation that my current results are flawed. Here is my sample test: I took the 100-day Previous-Day's-Close-to-Today's-High average overnight Run value of the stock whose symbol is PODD. That average overnight Run was +.38 (from April 11 to September 2, 2008). I then counted back from the Highs of days 101 to 105 and synchronized them with the Closes of days 102 to 106 and found that the average overnight Run from the Previous-Day's-Close-to-Today's-High for this 5-day stretch was exactly +.96. It added approximately +.03 to the 100-day average of +.38 bringing the 105-day average up to +.41 as my test filter verifies. It seems to me that StockFetcher was "able to increase that [100-day] value" based upon my small sample test. I believe that it is accurate up to 450-days, but will change my point of view if proven wrong. That's just my two-cents worth. Thanks Chetron for debugging my filter and thanks to TRO for the original inspiration for this filter. jhb2 P.S. I'm hunting for back info to solve the ACCURACY question. Here's one thread: http://forums.stockfetcher.com/sfforums/?q=view&fid=1001&tid=58590&qrid=&isiframe= |
| TheRumpledOne 6,411 posts msg #66899 - Ignore TheRumpledOne |
9/2/2008 4:08:37 PM /* 1 -- Master Run Prognosticator (Combined -- Previous-Close and Open-to-High-Today) */ Didn't I post something like this in Run Forest, Run? Previous Close to High and open to high?? What's different about your filter? |
| jimhbutton 104 posts msg #66900 - Ignore jimhbutton modified |
9/2/2008 4:27:32 PM Hi TRO, Thanks for the original ideas surrounding your Run-Pullback & Drop-Bounce filters. Personally, I think they're brilliant. I was in the middle of trading today and hit a glitch with my filter (which is nothing but your Run idea modified to fit my personality and trading style but titled to conform to my table of index). I was freaking out and didn't have time to give credit to whom credit is due; my money was on the line, my heart was in my throat, I was thinking "what the hell else did I screw up in my programming", and "are the stats that I entered the trade with flawed"? Chetron graciously debugged my programming and in the meantime got me thinking about the whole 100-day count limit thing. Thanks a million TRO for giving me the tools to fill my toolbox. You literally are the one who has made me a trader with the generous sharing of your secrets. jhb2 P.S. You can call me "TRO jr." or "son of TRO" if you want. |
| StockFetcher Forums · General Discussion · Debug Issue with: "Warning: period length exceeded." | << 1 2 >>Post Follow-up |
Copyright 2022 - Vestyl Software L.L.C.•Terms of Service | License | Questions or comments? Contact Us
EOD Data sources: DDFPlus & CSI Data
Quotes delayed during active market hours. Delay times are at least 15 mins for NASDAQ, 20 mins for NYSE and Amex. Delayed intraday data provided by DDFPlus