| SwingTimerTS2k Indicator Page: 15 | |
HOME WhySwT-a WhySwT-b WhySwT-c ======= Intro Page 1 Page 2 Page 3 Page 4 Page 5 ======= Params Page 6 Page 7 Page 8 Page 9 Page 10 Page 11 Page 12 Page 13 Page 14 Page 15 Page 16 Page 17 Page 18 Page 19 Page 20 Page 21 ======= |
Of great importance in the best analyses of datasets is having enough data available to be able to create a history that covers the types of markets we are experiencing today (yes, IF YOU HAVE ENOUGH DATA I GUARANTEE, it has happened before). The following pictures are designed to assist you in this understanding. We will be working with the Dow Jones Industrial data since relatively long sets of these data are available. In doing this study we installed two instances of STI. One uses a NBar value of 13 and the other has a NBar value of 34. Both use values for Groups of 4 and for MatchN of 1 (except as noted in the very last display). The top portion starts in January, 1970. It appears to have both 13 and 34 bar boxes associated with all the turns that were detected at these NBar lengths. Now look at the next picture which starts in January, 1980 -- 20 years of data -- with ten years less data suddenly we lose some 34 bar boxes and the existing 34 bar boxes have a different size. Remember, we are asking for AVERAGING OF 4 GROUPS for selecting statistics. Since under all such arrangements we discard 4 values, which are extreme measures, then we must have at least 8 occurrences of a particular swing from some given pattern number for such a computation to be made.
Now to add a couple more pictures.
The upper picture of this pair uses same parameters as the first two and moves the starting date to January, 1990. Now ALL of our 34 bar boxes have disappeared. REASON -- not enough data for swings of that length. The 13 bar boxes are "hanging in" but are changing location and size a bit because of the reduced data set. If you think about it, looking 13 days ahead ain't too bad !!! In the bottom picture the MatchN parameter was changed to ZERO (0). This means that for any upgoing projection we use ALL prior upgoing legs in the computation and for downgoing projections we use all the prior downgoing legs in the computation. Taking this approach we now have sufficient 34 bar data to begin again to make such projections. YOU MUST EXPERIMENT WITH THE NBAR, GROUPS, and MATCHN PARAMETERS AND MATCH THEM TO YOUR NEEDS AND DATA SET. CLICK THE Add to Cart Button TO ORDER the |
