Registered User Joined: 10/28/2009 Posts: 40

Is there a PCF for a fry pan or gradual rounded bottom such as LXK shown below.

Worden Trainer
Joined: 10/7/2004 Posts: 65,138

You may want to take a look at the PCF for rounded bottom trading topic.
Bruce Personal Criteria Formulas TC2000 Support Articles

Registered User Joined: 10/28/2009 Posts: 40

I read all posts and tried the PCFs. Nothing worked. Please write a PCF that will find patterns like LXK.

Worden Trainer
Joined: 10/7/2004 Posts: 65,138

To create a Condition Formula for a rounded bottom, we would need a specific unambiguous objective definition of the pattern you are attempting to identify. We can't work from ambiguous I know it when I see it type definitions. An image might be useful in clarifying an already good definition, but does not in and of itself constitute a definition.
Note that in the PCF for rounded bottom trading topic, we started with a specific definition of a rounded bottom over 10bars. When it did not produce the desired results, esc952 came up with specific requirements which could be implemented in a formula. The resulting formula identified symbols meeting those specific requirements.
If the formula did not work for you, it is because your definition of a rounded bottom is different than esc952's definition of a rounded bottom. This is not surprising.
In order to identify symbols that you see as a rounded bottom, you need to start by specifying the exact requirements of what constitutes your version of a rounded bottom. Then we can see if it is possible to identify those requirements using TC2000 (it may or may not be possible depending on the specific definition).
Let's start with something which is currently true for LXK which builds on the linear regression slope concept esc952 was exploring. The 20period linear regression slope for LXK has been increasing for all but 1 the past 30 bars (9/30/2013 was the exception).
We can create a Condition Formula for the regression slope increasing from one bar to the next as follows:
19 * (C + C20) > 38 * AVGC19.1
Using Linear Regression vs Classical Peaks/Valleys for Divergence Analysis
If we use that as the Boolean Formula in a Custom PCF % True Indicator with the Average Type set to Simple and the Period set to 30, that Custom PCF % True Indicator will return 96.67%.
Adding and Moving Indicators
So we could create a condition which currently returns for LXK by clicking on the Custom PCF % True Indicator and selecting Create Scan Condition to create a condition for the value of the Custom PCF % True Indicator being greater than 95.
Create Conditions from Your Chart
You could also just click on the Custom PCF % True Indicator and select Show Values in WatchList to create a WatchList Column which could be used as a sort.
Sorting with Indicators
You can adjust the period of the Custom PCF % True Indicator. We can create a different formula for a different linear regression slope period. We could compare the linear regression slope to 2 bars ago instead of the previous bar (which is currently true for 26 bars in a row):
19 * (C + C21) + 17 * (C1 + C20) > 72 * AVGC18.2
We could check for the linear regression slope to be negative at some point in the past. For example, the following checks for the 20period linear regression slope to have been negative 30 bars ago:
19 * (C30  C49) + 17 * (C31  C48) + 15 * (C32  C47) + 13 * (C33  C46) + 11 * (C34  C45) + 9 * (C35  C44) + 7 * (C36  C43) + 5 * (C37  C42) + 3 * (C38  C41) + C39 < C40
We could use a definition not involving linear regression slope at all, but we have to have a definition before trying to identify the pattern.
Bruce Personal Criteria Formulas TC2000 Support Articles
