| stockfetcher 980 posts
 msg #112556
 | 4/1/2013 11:12:14 AM 
 The new DATE([PERIOD],[INDICATOR]) function allows you to access past values of measures in new ways. In function, this feature works similar to the "days ago" phrase, as it modifies the desired measure; however, you can provide specific instructions on the actual offset of the days ago phrase.
 
 The example below returns matches that are more than 10% above the opening price for the year.
 
 
 
 
 To find stocks that are below the RSI(10) from the start of the month:
 
 
 
 
 Valid options for the "PERIOD" argument:
 
 
 
 YEARSTART: Initial value of the indicator for the year
 YEAREND: Last value of the indicator for the prior year
 MONTHSTART: Initial value of the indicator for the month
 MONTHEND: Final value of the indicator for the prior month
 MONDAY - FRIDAY: Value from the specific day of the week
 YYYYMMDD: Specfic date
 
 Here are a couple more examples:
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Please let us know if you run into any issues with this new function, have any question about how it works, or would like to see any modifications.
 
 StockFetcher Support
 
 
 
 | 
| mahkoh 1,065 posts
 msg #112563
 - Ignore mahkoh
 modified
 | 4/1/2013 4:13:54 PM 
 +1000 kudos to the SF crew!
 
 Edit: that is if I can get this to work
 
 
 
 This should give me lines at monthly pivot lines. It does not however, although check syntax OK's everything.
 Is there a way to code this?
 
 
 | 
| stockfetcher 980 posts
 msg #112566
 | 4/1/2013 4:51:24 PM 
 If possible, can you provide an example where the numbers do not appear to match up with the syntax in your filter? For example, if you can supply a symbol and expected value, that would help us track where there may be any issues.
 
 Thanks!
 
 StockFetcher Support
 
 
 
 
 | 
| mahkoh 1,065 posts
 msg #112567
 - Ignore mahkoh
 | 4/1/2013 4:57:52 PM 
 For some reason it didn't work when I made the changes in my original filter that I had to adjust for lookback days daily:
 
 set{mh,high 21 day high             20 days ago}
 set{ml,low 21 day low              20  days ago}
 set{cl,close             20  days ago}
 
 to
 
 set{mh,date(monthend,high 21 day high)}
 set{ml,date(monthend,low 21 day low)}
 set{cl,date(monthend,close)}
 
 but clicking my posted filter above works perfect. +1000 it is, thanks!
 
 
 | 
| olbn 4 posts
 msg #113880
 - Ignore olbn
 | 6/3/2013 6:19:50 PM 
 Is it possible to add for Optionable stocks to analyze from the Option Expiration Month (Generally 3rd Friday of the Month)?
 
 
 | 
| dangreene 229 posts
 msg #156922
 - Ignore dangreene
 | 6/4/2021 9:27:49 PM 
 POP to TOP
 
 
 | 
| styliten 332 posts
 msg #159584
 - Ignore styliten
 modified
 | 9/11/2022 9:18:47 PM 
 Here is another way of drawing RELATIVE reference price lines by month based on the entire price range from low 1 month low to high 1 month high:
 
 The following assumes 21 days in an average trading month; feel free to adjust [21] days to 15 or 30 or as you see fit.
 
 A volatile prior month will automatically make the reference price lines for the following month more widely spread out; conversely, a quiet month (in consolidation) will make reference price lines meshed up together. The underlying logic being what if the price trend in the immediate prior month were to continue exactly as it has been? As a result any price behavior pattern deviated from prior month will stand out very quickly to catch one's attention. Because these relative reference price lines stay constant for the entire duration of the current month, there is no moving averages or other curves to distract one's eyes.
 
 
 
 
 It can capture price shifts/reversals at bottom and tops better those those based off on "moving averages" or linear regression intercepts alone.
 
 
 | 
| TheRumpledOne 6,529 posts
 msg #161699
 - Ignore TheRumpledOne
 modified
 | 3/15/2025 12:38:43 PM 
 Is there a way to only use data from a certain day of the week?
 
 Example:
 
 I want to know how many Fridays closed above the open over the last 5 Fridays.
 
 
 
 
 The above doesn't return correct values.  Am I doing something wrong?
 
 Need a DAYTOUSE(FRIDAY) phrase so only FRIDAY data used in calculation.
 
 
 
 
 
 | 
| nibor100 1,099 posts
 msg #161704
 - Ignore nibor100
 | 3/16/2025 11:56:01 AM 
 I wouldn't say doing something wrong, just possibly forgetting that SF's count command parses on a single day basis so in your filter when it looks at the 5 most recent days it finds that the current Fri has var3d above 0 for NVDA and Thursday it finds that the prior Fri has var3d above 0, and so on for Wed, Tue, and Mon bringing the count total to 5.
 
 It is not going to count the last 5 Fri values as they are each 5 days apart and Count only works on a daily basis.
 
 If you add "draw var3d" to your filter you'll see what I mean graphically.
 
 The following will possibly get you close to what you want:
 Ed S.
 
 set{nFRI, count(var3d above 0, 1)}
 set{neFRI, sum(nfri, 25)}
 set{newFRI, nefri/5} draw newfri
 add column nFRI
 add column neFRI
 add column newFRI
 
 
 |