15minofPham
170 posts
msg #160112
- Ignore 15minofPham |
2/18/2023 1:24:09 PM
I've figured out how to calculate the avg high > open % for all weeks but is there a way to break it down to between up and down weeks?
|
nibor100
1,044 posts
msg #160114
- Ignore nibor100 |
2/18/2023 6:20:00 PM
Post your filter as it is so far and I'll work on what you now want.
Thanks,
Ed S.
|
15minofPham
170 posts
msg #160115
- Ignore 15minofPham |
2/19/2023 12:32:01 AM
Hi Ed,
Here it is. Thanks for looking at it!
set{wclop, weekly close - weekly close 1 week ago}
set{Long_Profit, wclop/ weekly close 1 week ago}
set{wkprofitpct, 100 * Long_Profit}
set{b1, volume - average volume(63)}
set{s1, b1 / average volume(63)}
set{abvavgvol63, s1 *100}
set{c1, weekly high-weekly open}
set{t1, c1/weekly open}
set{highabvopen%, t1 *100}
set{whiop2, weekly high - weekly open}
set{LongProfit2, whiop2/weekly open}
set{highabvopen2%, 100 * LongProfit2}
set{d1a, count(LongProfit2 between .08 and .10, 20)}
set{d2a, count(LongProfit2 between .10 and .15, 20)}
set{d3a, count(LongProfit2 between .15 and .20, 20)}
set{d4a, count(LongProfit2 > .20, 20)}
set{total1, count(LongProfit2 between .08 and .10, 20) + count(LongProfit2 between .10 and .15, 20)}
set{total2, total1 + count(LongProfit2 between .15 and .20, 20)}
set{total, total2 + count(LongProfit2 > .20, 20)}
set{total3, total/20}
set{total%, total3 * 100}
set{diff, close - low}
set{range, high - low}
set{pos, diff/range}
set{daypos, pos *100}
add column wkProfitPct
add column abvavgvol63
add column average day range(21)
add column daypos
add column separator
add column cma(highabvopen%,20)
add column count(highabvopen% above 3%, 100){wkly3%}
add column count(highabvopen% above 5%, 100){wkly5%}
add column separator
add column highabvopen%
add column weekly open
add column weekly high
add column separator
add column total%
add column total
add column d1a{8-10}
add column d2a{10-15}
add column d3a{15-20}
add column d4a{20+}
add column separator
add column count(weekly close > weekly close 1 week ago, 4){up/4}
add column count(weekly close > weekly close 1 week ago, 8){up/8}
add column count(weekly close > weekly close 1 week ago, 12){up/12}
add column count(weekly close > weekly close 1 week ago, 20){up/20}
add column separator
add column RSquared(close,10)
add column RSquared(close,21)
add column RSquared(close,63)
add column RSquared(close,100)
average day range(21) above 1.00
sort column 28 descending
|
nibor100
1,044 posts
msg #160116
- Ignore nibor100 |
2/19/2023 12:25:16 PM
At first glance, it looks like you can get the totals for down weeks by adding new "add column statements similar to your up4, up8, etc. and reversing the greater than sign.
or
you can make your up4, up8, etc. add column statements into set variables and then just subtract that variable from the counts 4,8,12,20 depending on how many weeks you are interested in covering.
I noticed you have weekly high vs weekly open in some places and weekly close vs prior weekly close in other places which is fine if that's what you intended.
If this doesn't work for you let me know and I'll work the actual filter lines after my grandson leaves town tomorrow.
Thanks,
Ed S.
|
15minofPham
170 posts
msg #160117
- Ignore 15minofPham |
2/20/2023 3:11:19 AM
I'm looking for the avg weekly high > open % on up vs down weeks. I'm not sure if SF has the ability to break it down to these two variables.
|
nibor100
1,044 posts
msg #160118
- Ignore nibor100 |
2/20/2023 11:57:04 AM
Just to double check what you are after:
an up week is defined by weekly close > weekly close 1 week ago
a down week is defined by weekly close < weekly close 1 week ago
so in a 5 week period with 2 down weeks and 3 up weeks
you want the average of the weekly high vs the weekly open for the 3 up weeks
and also the avg of the weekly high vs the weekly open for the 2 down weeks.
are my assumptions correct?
what do you want in the situation where a down week has the weekly high is the weekly open?
Thanks
Ed S.
|
15minofPham
170 posts
msg #160119
- Ignore 15minofPham |
2/20/2023 6:50:10 PM
You are correct Ed.
Regarding this: what do you want in the situation where a down week has the weekly high is the weekly open?
Please include it in the avg
|
nibor100
1,044 posts
msg #160120
- Ignore nibor100 |
2/21/2023 1:49:17 PM
This filter snippet should be what you are after.
Let me know if it is not.
Ed S.
|
15minofPham
170 posts
msg #160121
- Ignore 15minofPham |
2/21/2023 3:14:33 PM
Great job Ed!
The Up weeks calculations are very close to what I got, but the down weeks are way off so maybe I'm not reading your filter correctly? This is the high > open total I got for TSLA (not including this week) over the past 10 weeks.
Up Weeks
2/13: 11.95
2/6: 10.88
1/30: 11.77
1/23: 32.98
1/16: 8.74
1/19: 5.88
12/26: 5.94
Sum = 88.14
Avg = 12.59
Your totals
Sum = 78.89
Avg = 12.63
Down weeks
1/2: 0.28
12/19: 0.81
12/12: 1.27
Sum = 1.81
Avg = 0.60
Your totals
Sum = 12.63
Avg = 4.21
Thank you so much for helping me on this!
|
nibor100
1,044 posts
msg #160122
- Ignore nibor100 |
2/21/2023 11:57:42 PM
When I run the posted filter on Friday 2/17, I get the same results as you have listed for the up weeks and down weeks totals and averages.
When I track thru the past week by week I get the individual date values for both up weeks and down weeks that you have listed, though many of the dates are different.
for example you have 2 consecutive up weeks occurring in just 3 days apart 1/19 and 1/16.
And I don't find any of the values you have listed in either of the "your totals" sections
I do notice that the total of 12.63 for down weeks is identical to avg of 12.63, shown for up weeks, perhaps a typo?
I'll check again in the morning as some data providers do adjustments long after the market closes.
Ed S.
|