Refresh of data and result is completely different. I suppose the NSF doesn't intervene in this.
Rename
Number of occurrences doesn't match.
First chart 57 and second 78,

First chart 57 and second 78,
I took a careful look at the bottom of your chart (3rd screenshot) and noticed that its data range is set to 3 years. The Evolver is configured to run on 500 bars and 2 years on the other two screenshots. This difference explains what you're seeing. Make them match and it should disappear.
And, since this question was somehow asked in "Candlestick Genetic Evolver: chart shows SPY, not symbol" but presumably did not have to do with that issue, I've extracted it into a new topic for the forum to stay ordered.
And, since this question was somehow asked in "Candlestick Genetic Evolver: chart shows SPY, not symbol" but presumably did not have to do with that issue, I've extracted it into a new topic for the forum to stay ordered.
Why do the number of observations change when i adjust the N Bar return?
I don' t see how a return can influence a number of observations?



I don' t see how a return can influence a number of observations?
A larger N Bar return requires more data, so therefore more lead bars, so fewer observations
I stil don' t get it.
My way of thinking.
You have a bullish long white line. And then you start to calculate what's the return for N days,
But i suppose my logic is not correct,
My way of thinking.
You have a bullish long white line. And then you start to calculate what's the return for N days,
But i suppose my logic is not correct,
Patterns that occur in the final 10 bars of data will not be counted because the return cannot be calculated. Sorry for any confusion.
Glitch explained in a previous post that last 10 days doesn' count. So we can skip 1 occurence. There are 3 occurences left that i can't explain.

Most recent 1 year and chart goes until 27 september. Date in the future.
I would guess you added future bars to the chart by dragging the horizontal margin line along the right axis. The scroll bar thumb position supports this idea.
I'm a right if a only want to see the number of observations i have to take 1 N-bar return.
Your Response
Post
Edit Post
Login is required