indexa
32 posts
msg #110162
- Ignore indexa |
1/14/2013 7:42:39 PM
1/14/2013 7:29:28 PM
Hi,
Would it be possible to add Cutler's RSI indicator to Stockfetcher library?
Or, is there a way to calculate this indicator:
A variation called Cutler's RSI is based on a simple moving average of U and D, instead of the exponential average above
RSI = 100 - 100 / ( 1 + RS)
where
RS = SMA(U,n) / SMA(D,n)
Thanks!
|
evo34
82 posts
msg #110213
- Ignore evo34 |
1/15/2013 5:26:40 PM
I second this..
|
Kevin_in_GA
4,599 posts
msg #110218
- Ignore Kevin_in_GA |
1/15/2013 9:42:38 PM
Why would this be any better? The use of one MA type versus another usually results in only small variations in the resulting number. Can you show an example of how it is different/better? I'm honestly curious.
Kevin
|
indexa
32 posts
msg #110221
- Ignore indexa |
1/16/2013 12:01:51 AM
Sure, Kevin, I will try to explain.
I learned about RSI(2) in this forum and now I use it to spot my entry.
An inflexion in this indicator shows me the reversal trend.
Suppose a stocks shows, for a moment, a red candle and maybe the end of a rally. The traditional RSI (with EMA) shows a negative line; but, the Cutler's RSI slope is above 0. Later, the candle shifts and it's green. The Wilder RSI now is not negative (maybe positive or = 0). The difference is Cutler's RSI remains positive and there is no false trigger.
Regards.
|
guru_trader
485 posts
msg #110292
- Ignore guru_trader
modified |
1/17/2013 12:27:57 PM
Here is my attempt at Cutler's RSI. I noticed there are several definitions of how to calculate it, so let me know if it should be calculated differently. To change the lookback days, simply change the "14" to whatever days you want.
* fixed code to account for situation when RS denominator (sum of down days) = 0
|
10 posts
msg #110297
- Ignore |
1/17/2013 1:51:09 PM
Very impressive!
Unfortunately it doesn't run with the SF standart subscription.
Thanks anyway.
|
guru_trader
485 posts
msg #110298
- Ignore guru_trader
modified |
1/17/2013 2:16:26 PM
Then it's time to upgrade your subscription! :)
Actually, I just caught an error in the script. I didn't consider when the denominator is 0 ... I'll fix that soon. [*fixed, and posted above]
|
indexa
32 posts
msg #110379
- Ignore indexa |
1/21/2013 9:40:05 AM
Yes, you're right, guru!
Thanks
|
duke56468
683 posts
msg #110380
- Ignore duke56468 |
1/21/2013 10:17:30 AM
Can someone explain what constitutes "performance restrictions"? Some filters seem very long and are no problem for us in the cheap seats, and others like this one which appears short wont run. Is it the number of set definitions?
|
Kevin_in_GA
4,599 posts
msg #110384
- Ignore Kevin_in_GA |
1/21/2013 12:23:25 PM
I usually run into that issue when nesting user defined variables (e.g., create a user-defined variable, subtract another user-defined variable from it, then make a custom MA from that difference, then ...)
After a few it hits the wall.
|