styliten
278 posts
msg #153634
- Ignore styliten |
8/19/2020 12:21:02 PM
To me, MONTHLY chart/filter would top everything else. Is adding the monthly chart/filter really really that difficult to implement?
and KAMA would be the distant second.
Anyone else?
|
xarlor
562 posts
msg #153641
- Ignore xarlor |
8/19/2020 8:28:16 PM
1. Reimplement backtesting functionality. I would pay extra for the feature.
2. If/Then, While/Do, For loops.
3.. Drop a date or DaysAgo into a variable for recall later.
4. More indicators. Always more indicators.
|
shillllihs
5,974 posts
msg #153643
- Ignore shillllihs |
8/19/2020 8:56:45 PM
3 more inches.
|
ron22
255 posts
msg #153651
- Ignore ron22 |
8/19/2020 11:01:02 PM
add: VWAP: Volume Weighted Average Price
|
styliten
278 posts
msg #153654
- Ignore styliten |
8/20/2020 7:30:38 AM
SF is obviously capable of doing IF/THEN:
https://www.stockfetcher.com/forums/Indicators/Williams-Accumulation-Distribution/76455
but not granted to end user?
|
graftonian
1,089 posts
msg #153656
- Ignore graftonian |
8/20/2020 9:05:31 AM
What I miss is:
The ability to use the result of one variable in another function.
eg:
set{var1, xyz}
set{var2, DATE(xyz, close)}
|
nibor100
1,010 posts
msg #153663
- Ignore nibor100 |
8/20/2020 4:11:55 PM
@styliten,
Regarding monthly charts/data/filters it is probably a limitation of the data provider SF uses and anyway do you really want a 2 year chart with only 24 candlesticks, as SF seems to limit the max candlesticks displayed to what fits in a 2 year window.
There is an involved work-a-round for using IF Then Else in filters and passing that kind of programming capability into the SF english phrase interface with error checking is probably a bit harder than we might think it is.
@xarlor,
I wasn't a SF member back when they had backtesting but from the few searches I've performed it seemed no one was happy with it and there were many problems and lots of SF help requests which is probably why they dropped it.
@graftonian
SF already has that capability to some degree, but your example seemed a little off based on the point I thought you were making.
My 2 wishlist items are:
1. a much better filter debugger/error messages/syntax checking etc.
2. Add the Help message repository to the Forums.
Ed S..
|
fotchstecker
304 posts
msg #154140
- Ignore fotchstecker |
9/30/2020 9:57:35 PM
Improved backtesting, no question. Backtesting has its own problems, but if your assumption is that filters are fitted then you can deal with those problems. I just want more stats.
|
styliten
278 posts
msg #154199
- Ignore styliten
modified |
10/6/2020 1:43:20 PM
Fell in love with Center Standard Error Bands indicator:
https://www.stockfetcher.com/forums/Indicators/Standard-Error-Bands/96315
This is in fact the LRI envelope that Kevin lamented about lacking:
https://www.stockfetcher.com/forums/Filter-Exchange/Any-chance-of-seeing-a-LRI-Envelope/108033/-1/108037
It is, after all, NOT lacking, but under a totally different name!
The only problem with Center Standard Error Bands at this time is that the MAXIMUM period allowed is 252, unlike LRI that allows maximum period of 300! According to some, the LRI(100) and LRI(300) have special importance in FX trading (on the hourly chart). This discrepancy between LRI and Center Standard Error Bands on Maximum period has been communicated to SF developers numerous times and at this very moment, they are taking another look and see if maximum period for Center Standard Error Bands can be easily upgraded to 300 as well.
If you like CSEB after reading this post, let's beg for Standard Error Band(300)! Have some mercy on us, please!
Sample Code:
|
marine2
963 posts
msg #154207
- Ignore marine2 |
10/7/2020 2:22:29 PM
Bringing back backtesting. The last version SF had once you learned its nuances it really helped you in your filter development and whether to trust that filter or not. Of course you needed to test your filter against a oversold market, undersold market, and a neutral market environment measuring of course against the S&P 500, and Nasdaq markets.
The main reason why SF stopped its backtesting tool was it cost too much for them to utilize. Which I thought was silly since all they needed to do is charge more on their basic subscription and for their advanced. I think the subscribers that didn’t take the time to learn how to use it effectively probably irritated the SF staff to the point they said ok then, we will just discontinue providing it. Not the solution of course to fixing their problem but instead SF should have been more patient with their customer or found ways to correct the complexity of it.
A good analogy of this would be, you have a car dealership that sells you that product but, you provide no service for it. When you have a service that allows you to create and use a filter you should then have a testing mechanism that also allows you to backtest your creation. SF gives you half a tool.
Hopefully in the near future they will bring back their backtesting tool?
|