I can calculate the 14-day moving linear regression of the 28-period simple moving average of the close using the least squares method as follows.

AVG(AVGC28,14) + 6.5 * (6.5 * AVGC28 + 5.5 * AVGC28.1 + 4.5 * AVGC28.2 + 3.5 * AVGC28.3 + 2.5 * AVGC28.4 + 1.5 * AVGC28.5 + .5 * AVGC28.6 - .5 * AVGC28.7 - 1.5 * AVGC28.8 - 2.5 * AVGC28.9 - 3.5 * AVGC28.10 - 4.5 * AVGC28.11 - 5.5 * AVGC28.12 - 6.5 * AVGC28.13) / 227.5

But a Median-Median or Triple Median Linear Regression results in a formula which is way to long to be practical or post in the forums.

If the 14-day moving linear regression of the 28-period simple moving average of the close is good enough, then you can use it as the Formula in a Custom PCF Indicator. This would allow you to set the Plot Style and colors of the Custom PCF Indicator as desired. You could then add a moving average to the Custom PCF Indicator and adjust its settings as desired as well.

Adding and Moving Indicators (7:19)

AVG(AVGC28,14) + 6.5 * (6.5 * AVGC28. + 5.5 * AVGC28.1 + 4.5 * AVGC28.2 + 3.5 * AVGC28.3 + 2.5 * AVGC28.4 + 1.5 * AVGC28.5 + .5 * AVGC28.6 - .5 * AVGC28.7 - 1.5 * AVGC28.8 - 2.5 * AVGC28.9 - 3.5 * AVGC28.10 - 4.5 * AVGC28.11 - 5.5 * AVGC28.12 - 6.5 * AVGC28.13) / 227.5

 bRUCE COPY PASTE IN TC200 AND DONT WORK AND ALL THE INDICATOR  AND EXP ATR EITHER WAY ARE WRONG 

LIKE YOU MAKE THE CORRECTION 

THANKS BRUCE

LOOKINF FORWARD FOR AN EXPLANATION

If there is not an option to paste available when right-clicking in the window where you wish to use the formula, you should be able to use ctrl-v to paste instead.

One issue you might experience when using formulas outside of the United States is if your decimal marker is set to use a , instead of set to use a . as would be common in the United States. The only good way around this issue if you don't want to have the decimal markers set to . on your computer would be to re-write any formulas to avoid the use of decimal markers.

For example, the formula given above could be re-written as follows to eliminate the decimal markers.

AVG(AVGC28,14) + 13 * (13 * (AVGC28 - AVGC28.13) + 11 * (AVGC28.1 - AVGC28.12) + 9 * (AVGC28.2 - AVGC28.11) + 7 * (AVGC28.3 - AVGC28.10) + 5 * (AVGC28.4 - AVGC28.9) + 3 * (AVGC28.5 - AVGC28.8) + AVGC28.6 - AVGC28.7) / 910

This was done by multiplying all of the coefficients by 2 and then multiplier denominator by 4 (because of the way the parentheses are placed the coefficients interact in such a way that 2 * 2 = 4).

I don't know what you are saying about "all the indicator and exp ATR either way are wrong."