StockFetcher Forums · Filter Exchange · When was the last time user request was actually approved/implemented on SF<< 1 2 3 >>Post Follow-up
styliten
277 posts
msg #152427
Ignore styliten
modified
6/3/2020 12:52:57 PM

Looking at the main indicator list page:

https://www.stockfetcher.com/forums/Indicators

It's pretty obvious that there are at least 2 misspellings in the indicator name:

a. Directional Movement Indcators +DI, -DI, ADX 8/10/2005 7:05

b. Positive Volme Index (PVI) 8/17/2004 14:39

while the misspellings are not critical to the utility of the underlying indicator, it makes one wonder whether these misspellings have never been pointed out by any user since their introductions on SF, which were 15~16 years ago.

If these are minor issues, what about "CUSTOM LINEAR REGRESSION INDICATORS"? They are obviously there and made available in April 2016:

https://www.stockfetcher.com/forums/Filter-Exchange/CUSTOM-LINEAR-REGRESSION-INDICATORS/127799/-1/127887

But "CUSTOM LINEAR REGRESSION INDICATORS" is still not officially listed under the main "indicator" page.

Lastly, has anyone ever had any luck getting their request for indicator addition/modification/enhancement actually approved? And when was the last time one's reasonable wish is granted? The product development team on SF seems so elusive with so few pronouncements that it's next to impossible to know what's really on the product development/enhancement team's mind. At certain time, the notion that product development/enhancement has abandoned SF has certainly crossed one's mind, which is spine chilling at the very best.



Cheese
1,374 posts
msg #152430
Ignore Cheese
6/3/2020 6:09:27 PM

@styliten,

Have you already used all sF tools and still not been able to achieve the results that you want?

If so, then you may want to post your best and most comprehensive filters, and ask the many
traders and coders here for suggestions. A few people always seem happy to help.

Best wishes.

styliten
277 posts
msg #152434
Ignore styliten
modified
6/3/2020 9:19:06 PM

Thank you for your kindness.

So far, I've only requested sporadic enhancements. Here are a few examples:

1. On the "settings" ==> Results tab, currently the maximum # of tickers PER PAGE is 100. I've requested new maximums of 250 and 500 be added, only because often times, my filter yields 300-600 tickers per day. This is easy to see, now that the market is definitely in the recovery mode, so a LOT of tickers are recovering fast, gaining 4~5% or more each trading day. Since I don't want to miss ANYTHING, I'd like to review them ALL (i.e., anything and everything that moved up more than 4 or 4.5% during the day, provided it meet very few other constraints, such as avg volume, or market cap). Having the maximum increased to 250 or 500 meant I would have far fewer pages to turn and thus increase the review efficiency tremendously.

If SF product team really have given this request genuine consideration and decided that adding 2 extra maximum #s is too much, they can simply choose to add one option of "show all" instead. At any rate, is a request this simple too difficult to implement in year 2020? Is this request unreasonable to begin with? Why hasn't there been any official response from the SF's development team at all? Is there still an active product development team in existence as of today? Hence my growing desire to know what kind of service experience other active users have been getting from SF's product development team lately.

2. I've been relying more and more on linear regression related indicators, and in particular, center standard error band. True, there is already lri() indicator, but each LRI() only draws a thin line on the chart whereas center standard error band actually draws a band that allows user to control band width appearance, much like ma/ema envelope to simple ma/ema.

Then I realized the built-in band width for center standard error band only allows 1 decimal place, unlike any other linear regression related indicators which allow 2 decimals. To see why the # of decimal places actually do matter, simply try this code:

============================
date offset is 1

dow 30

add column stderr(close, 5) {std5}
add column stderr(close, 50) {std50}
add column stderr(close, 50) divided by stderr(close, 5) {std50vs5}
sort on column 7 descending

============================

so as of yesterday (6/2/2020), the std err (50) for UNH is 60+ times larger than its std err(5)! This means if you try to add

draw center standard error band(5, 0.1) [0.1 is simply the smallest value that will actually be used by this particular indicator to control the band width: any input values between 0.05 and 0.14 is rounded to 0.1, and anything less than 0.05 is treated as if 0]

draw center standard error band(50, 0.1)

on the chart, the latter would appear 60+ times WIDER than the former, an awkward looking band that too often overlaps with other add-on measures (such as center standard error band(25, 0.1)) during bottoming out period when just about all the indicators are trying to cross over each other in rapid succession.

Calculation shows band width needs to be somewhere in the 0.01~0.02 range in order for center standard error band(50) or larger not to cut through a broad swath of candlesticks on the chart; of course 3 decimals give user more wriggle room to play with the more exact band width so that all center standard error bands regardless the periods can appear to have approximately the same width and each crossover touches nothing but only 1 (or at most 2) candlesticks. This way the exact timing can be effortlessly determined within a second or less by just looking at the price chart itself! (I have hundreds of individual charts to look at each and every time, and multiple times per trading day; so every ease, every convenience, every second saved per chart per review counts and the savings do add up quickly to say the least)

Since the ultimate goal of having charts is to aid quick review and decision making process in the first place, I've asked, entreated and begged SF to increase the band width from 1 decimal place to at least 2 (3 is MUCH preferred).

This request has gone NO WHERE for at least the past 6 months! Hence my frustration, because this happens to be my all time favorite indicator for the main price chart and its efficacy puts just about any other measure in shame.

Cheese
1,374 posts
msg #152440
Ignore Cheese
6/4/2020 11:18:37 AM

styliten,
I don't have any answers or solutions to the issues you've highlighted.
I hope that some techies like xarlor, Ed S, or V E may have workarounds for you.

Cheese
1,374 posts
msg #152442
Ignore Cheese
modified
6/4/2020 11:30:11 AM

TRO TheRumpledOne popped in last month, but has remained largely silent.
snappyfrog, davesaint86 and safetrade are also very clever.
four is also good but seems to have left the building
Members like TRO, snappyfrog, davesaint86, safetrade and four are very good at techie stuff
so I hope that someone may be able to help you with a workaround.

styliten
277 posts
msg #152445
Ignore styliten
6/4/2020 11:51:52 AM

Thank you again, Cheese!

I really missed Kevin_in_GA. He is not only knowledgeable and nice, but also seemed to be the few heavy users that actually did pushed through SUCCESSFULLY with the SF development team, particularly Tomb.

There are a few other smaller grievances, such as the missing function of average(x, y). Yes, one surely can declare 2 variables using SET{} for every average one wants to calculate, but why NOT add the simple function of average/avg to begin with?

I spent the first 500 to 800 hrs as a newbie like everyone else, trying to find that holy grail of killer query that can do everything; but after that, I became fully convinced that such holy grail may not exist after all since every breakout has its own uniqueness or comes with different size shape or manner so much so that any predefined mold wouldn't fit most of these anyway. So instead, I've chosen to live like a gardener, taking out the tape measure and measure every single plant in the garden so that every single plant that has grown at least 2 inches overnight (up 4~4.5% today) will for sure be looked at or checked out. This way, the most promising plants will be given prompt and special attention asap.

My idea of checking them all on the daily basis was also boosted by the fact that that summary charts are indeed available on stockcharts.com:

https://stockcharts.com/freecharts/candleglance.html?EFA,EEM,EWA,EWO,EWK,EWZ,EWC,FXI,EWQ,EWG,EWH,INDA,EWI,EWJ,EWM,EWW,EWN,RSX,EWS,EWY,EWP,EWD,EWL,EWT,EWU|B|0

where one can view 25 tickers at once. Further, one can add price comparison to S&P 500 or Nasdaq 100 as a subgraph:

https://stockcharts.com/freecharts/candleglance.html?EFA,EEM,EWA,EWO,EWK,EWZ,EWC,FXI,EWQ,EWG,EWH,INDA,EWI,EWJ,EWM,EWW,EWN,RSX,EWS,EWY,EWP,EWD,EWL,EWT,EWU|B|J[$SPX]|0

or

https://stockcharts.com/freecharts/candleglance.html?EFA,EEM,EWA,EWO,EWK,EWZ,EWC,FXI,EWQ,EWG,EWH,INDA,EWI,EWJ,EWM,EWW,EWN,RSX,EWS,EWY,EWP,EWD,EWL,EWT,EWU|B|J[$COMPQ]|0

Chart reading can not be much easier than that. But if one has determined to make a living out of stock trading, how could one afford NOT to take a daily look at ALL the stocks that are growing by the day (multiple times during the trading hours)? There is a certain routine/tedium element in every job on planet earth, isn't it?

Cheese
1,374 posts
msg #152448
Ignore Cheese
6/4/2020 1:05:57 PM

Thank you very much, styliten.
Your CandleGlance charts are excellent !

I've also set up some CandleGlance charts to help me follow kossvet's medical picks
and to get a feel for other opportunities not taken

batch #! APOP,APTO,AQST,AXSM,CLSD,CLSK,CODX,CYCN,CYDY

https://stockcharts.com/freecharts/candleglance.html?APOP,APTO,AQST,AXSM,CLSD,CLSK,CODX,CYCN,CYDY,SPY,QQQ,XLV,IBB,XBI|C|0

https://stockcharts.com/freecharts
/candleglance.html?APOP:SPY,APTO:SPY,AQST:SPY,AXSM:SPY,CLSD:SPY,CLSK:SPY,CODX:SPY,CYCN:SPY,CYDY:SPY,QQQ:SPY,XLV:SPY,IBB:SPY,XBI:SPY|C|0

https://stockcharts.com/freecharts/candleglance.html?APOP:IBB,APTO:IBB,AQST:IBB,AXSM:IBB,CLSD:IBB,CLSK:IBB,CODX:IBB,CYCN:IBB,CYDY:IBB,SPY:IBB,QQQ:IBB,XLV:IBB,IBB,XBI:IBB|C|0



batch #2 GNPX,SAVA,SNGX,TNXP,TORC,TRIL,UAVS,VXRT

https://stockcharts.com/freecharts/candleglance.html?GNPX,SAVA,SNGX,TORC,TRIL,UAVS,VXRT,SPY,QQQ,XLV,IBB,XBI|C|0

https://stockcharts.com/freecharts/candleglance.html?GNPX:SPY,SAVA:SPY,SNGX:SPY,TNXP:SPY,TORC:SPY,TRIL:SPY,UAVS:SPY,VXRT:SPY,SPY,XLV:SPY,IBB:SPY,XBI:SPY|C|0

https://stockcharts.com/freecharts/candleglance.html?GNPX:IBB,SAVA:IBB,SNGX:IBB,TNXP:IBB,TORC:IBB,TRIL:IBB,UAVS:IBB,VXRT:IBB,SPY:IBB,QQQ:IBB,XLV:IBB,IBB,XBI:IBB|C|0

The learning for me here is that people like kossvet can succeed using their niche expertise
flying in the face of technical analysis.

Happy trading

styliten
277 posts
msg #152449
Ignore styliten
modified
6/4/2020 1:31:12 PM

I saw you like to see the price moves against S&P 500. I do too.

Again here is how I normally see the comparison using Dow 30 as an example:


dow 30
chart-time is 15 days

and draw Center Standard Error Band(10, 0.1)
and draw Center Standard Error Band(25, 0.1)
and draw Center Standard Error Band(45, 0.1)

and draw cema(relative strength(spy, 20), 9) line at 1
and draw cema(relative strength(spy, 50), 9) line at 1

[the reason I use spy instead ^SPX is that ^SPX (or any other indices) is only end of day, whereas SPY is available during the trading hours on any trading day; and for all practical purposes, the two can be treated as if identical to each other]

(Hopefully you'll see why I need 2 decimal places for bandwidth adjustment by now. And if you really look hard, did you see that PFE is on the verge of making a come back? If so, what kind of killer filter/algorithm can detect this kind of earliest telltale sign? Another example, see how early the warning signs actually showed up for UNH's downfall prior to today 6/4/2020.)

Cheese
1,374 posts
msg #152450
Ignore Cheese
modified
6/4/2020 2:50:42 PM

styliten,

I hope I understand your post correctly.

I was able to capture the minute differences.StockFetcher may be more versatile than you think.
Please look at the two charts in larger display, using the price candles or OHLC to see the differences.


Fetcher[
SYMLIST(PFE)
chart-time is 15 days
DRAW Upper Standard Error Band(5,0.1)
]




Cheese
1,374 posts
msg #152451
Ignore Cheese
modified
6/4/2020 2:53:18 PM

delete and re-post

StockFetcher Forums · Filter Exchange · When was the last time user request was actually approved/implemented on SF<< 1 2 3 >>Post Follow-up

*** Disclaimer *** StockFetcher.com does not endorse or suggest any of the securities which are returned in any of the searches or filters. They are provided purely for informational and research purposes. StockFetcher.com does not recommend particular securities. StockFetcher.com, Vestyl Software, L.L.C. and involved content providers shall not be liable for any errors or delays in the content, or for any actions taken based on the content.


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


This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.