FOLD Education

[MrBruceKA]

I would use a FOR. . . NEXT or DO . . . WHILE statement, if I was writing in BASIC programming language. However, I can only find a FOLD statement in thinkScript to iterate through statements.

Could someone convert these SCAN statements into a FOLD statement for me. BUT, more important, please tell me, in detail, what is going on. I can't understand the example giving in the TOS Help.

Thanks for the education

# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
input PSprd = 3;
def Pcnt = PSprd/100;
input RunLen = 8;

# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
# Get the average of the midpoints
def MidPoint = (Open+Close)/2;
def RunAvg = Average(MidPoint,RunLen);

# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
# Calculate the Percentages
def P1 = AbsValue((MidPoint[(RunLen-1)]-RunAvg)/RunAvg);
def P2 = AbsValue((MidPoint[(RunLen-2)]-RunAvg)/RunAvg);
def P3 = AbsValue((MidPoint[(RunLen-3)]-RunAvg)/RunAvg);
def P4 = AbsValue((MidPoint[(RunLen-4)]-RunAvg)/RunAvg);
def P5 = AbsValue((MidPoint[(RunLen-5)]-RunAvg)/RunAvg);
def P6 = AbsValue((MidPoint[(RunLen-6)]-RunAvg)/RunAvg);
def P7 = AbsValue((MidPoint[(RunLen-7)]-RunAvg)/RunAvg);
def P8 = AbsValue((MidPoint[(RunLen-8)]-RunAvg)/RunAvg);

# See if all are below the PSprd value
Def Ans = P1<Pcnt And P2<Pcnt And P3<Pcnt And P4<Pcnt And P5<Pcnt And P6<Pcnt And P7<Pcnt And P8<Pcnt;

 

[armybender]

Hi @MrBruceKA,

Here's an example of how you could do this using fold. One thing to understand is that fold can be used in a variety of ways, so for the purpose you've outlined the process is rather simple, but just to make sure it's stated, it can be used in more powerful ways, but it is not as advanced as a for/next or do/while loop in a traditional programming language.

Here's the code - I'll explain it below.

declare lower;            # Made this a lower study for testing purposes

# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
input PSprd = 3;
def Pcnt = PSprd / 100;
input RunLen = 8;

# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
# Get the average of the midpoints
def MidPoint = MidBodyVal();            # Simplified to use MidBodyVal() instead of (open + close) / 2
def RunAvg = Average(MidPoint, RunLen);

# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
# Calculate the Percentages
def P1 = AbsValue((MidPoint[(RunLen - 1)] - RunAvg) / RunAvg);
def P2 = AbsValue((MidPoint[(RunLen - 2)] - RunAvg) / RunAvg);
def P3 = AbsValue((MidPoint[(RunLen - 3)] - RunAvg) / RunAvg);
def P4 = AbsValue((MidPoint[(RunLen - 4)] - RunAvg) / RunAvg);
def P5 = AbsValue((MidPoint[(RunLen - 5)] - RunAvg) / RunAvg);
def P6 = AbsValue((MidPoint[(RunLen - 6)] - RunAvg) / RunAvg);
def P7 = AbsValue((MidPoint[(RunLen - 7)] - RunAvg) / RunAvg);
def P8 = AbsValue((MidPoint[(RunLen - 8)] - RunAvg) / RunAvg);

# See if all are below the PSprd value
def Ans = P1 < Pcnt and P2 < Pcnt and P3 < Pcnt and P4 < Pcnt and P5 < Pcnt and P6 < Pcnt and P7 < Pcnt and P8 < Pcnt;

# New fold answer
def foldAns =
    fold i = 0 to RunLen
    with x = 0
    do
    x + (AbsValue((GetValue(MidPoint, i) - RunAvg) / RunAvg) < Pcnt);


# Plots to test calculations
plot test = (P1 < Pcnt) + (P2 < Pcnt) + (P3 < Pcnt) + (P4 < Pcnt) + (P5 < Pcnt) + (P6 < Pcnt) + (P7 < Pcnt) + (P8 < Pcnt);
test.SetDefaultColor(Color.YELLOW);

plot testFoldAns = foldAns == RunLen;
testFoldAns.SetDefaultColor(Color.BLUE);

plot oldAns = Ans;
oldAns.SetDefaultColor(Color.WHITE);

plot newAns = foldAns;
newAns.SetDefaultColor(Color.CYAN);

First, I modified your MidPoint variable to use MidBodyVal() as a function. This is a built-in TOS function that calculates (open + close) / 2.

Next, the foldAns variable I created does exactly the same thing as your P1...P8 variables do.

Fold i = 0 to RunLen i is like saying "For i = 0 to RunLen". It's identical in that way.

With x = 0 simply defines the starting value for x as 0. This is one of the ways fold can vary - it is not always necessary to specify a starting value, but in this case it is since we're going to increment x if it meets a condition.

Do x + (AbsValue.....) adds the prior value of x plus 1 since the AbsValue argument evaluates to 1 or 0, depending on whether it is true or false. It's like saying 'if condition = true then x = x + 1 else x".

As a final step, to evaluate the statement in the way you evaluate Ans, I have a line testFoldAns below that simply checks if foldAns = RunLen. In your test variable you check all 8 conditions and it evaluates TRUE if all 8 variables are TRUE. In my version, I check if x has reached 8 (RunLen).

I hope that helps. Fold is tricky to understand, but once you get it then it's pretty easy to use.

Let me know if you have any further questions.

Thanks,

Adam

 

[MrBruceKA]

Hi @MrBruceKA,

Here's an example of how you could do this using fold. One thing to understand is that fold can be used in a variety of ways, so for the purpose you've outlined the process is rather simple, but just to make sure it's stated, it can be used in more powerful ways, but it is not as advanced as a for/next or do/while loop in a traditional programming language.

Here's the code - I'll explain it below.

declare lower;            # Made this a lower study for testing purposes

# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
input PSprd = 3;
def Pcnt = PSprd / 100;
input RunLen = 8;

# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
# Get the average of the midpoints
def MidPoint = MidBodyVal();            # Simplified to use MidBodyVal() instead of (open + close) / 2
def RunAvg = Average(MidPoint, RunLen);

# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
# Calculate the Percentages
def P1 = AbsValue((MidPoint[(RunLen - 1)] - RunAvg) / RunAvg);
def P2 = AbsValue((MidPoint[(RunLen - 2)] - RunAvg) / RunAvg);
def P3 = AbsValue((MidPoint[(RunLen - 3)] - RunAvg) / RunAvg);
def P4 = AbsValue((MidPoint[(RunLen - 4)] - RunAvg) / RunAvg);
def P5 = AbsValue((MidPoint[(RunLen - 5)] - RunAvg) / RunAvg);
def P6 = AbsValue((MidPoint[(RunLen - 6)] - RunAvg) / RunAvg);
def P7 = AbsValue((MidPoint[(RunLen - 7)] - RunAvg) / RunAvg);
def P8 = AbsValue((MidPoint[(RunLen - 8)] - RunAvg) / RunAvg);

# See if all are below the PSprd value
def Ans = P1 < Pcnt and P2 < Pcnt and P3 < Pcnt and P4 < Pcnt and P5 < Pcnt and P6 < Pcnt and P7 < Pcnt and P8 < Pcnt;

# New fold answer
def foldAns =
    fold i = 0 to RunLen
    with x = 0
    do
    x + (AbsValue((GetValue(MidPoint, i) - RunAvg) / RunAvg) < Pcnt);


# Plots to test calculations
plot test = (P1 < Pcnt) + (P2 < Pcnt) + (P3 < Pcnt) + (P4 < Pcnt) + (P5 < Pcnt) + (P6 < Pcnt) + (P7 < Pcnt) + (P8 < Pcnt);
test.SetDefaultColor(Color.YELLOW);

plot testFoldAns = foldAns == RunLen;
testFoldAns.SetDefaultColor(Color.BLUE);

plot oldAns = Ans;
oldAns.SetDefaultColor(Color.WHITE);

plot newAns = foldAns;
newAns.SetDefaultColor(Color.CYAN);

First, I modified your MidPoint variable to use MidBodyVal() as a function. This is a built-in TOS function that calculates (open + close) / 2.

Next, the foldAns variable I created does exactly the same thing as your P1...P8 variables do.

Fold i = 0 to RunLen i is like saying "For i = 0 to RunLen". It's identical in that way.

With x = 0 simply defines the starting value for x as 0. This is one of the ways fold can vary - it is not always necessary to specify a starting value, but in this case it is since we're going to increment x if it meets a condition.

Do x + (AbsValue.....) adds the prior value of x plus 1 since the AbsValue argument evaluates to 1 or 0, depending on whether it is true or false. It's like saying 'if condition = true then x = x + 1 else x".

As a final step, to evaluate the statement in the way you evaluate Ans, I have a line testFoldAns below that simply checks if foldAns = RunLen. In your test variable you check all 8 conditions and it evaluates TRUE if all 8 variables are TRUE. In my version, I check if x has reached 8 (RunLen).

I hope that helps. Fold is tricky to understand, but once you get it then it's pretty easy to use.

Let me know if you have any further questions.

Thanks,

Adam

Adam:

You must be a teacher because your explanation is exactly what I was looking for to understand the FOLD statement.

I'll take your code and play with it for a while to make sure I understand what's happening.

Thanks, Bruce

 

[armybender]

Adam:

You must be a teacher because your explanation is exactly what I was looking for to understand the FOLD statement.

I'll take your code and play with it for a while to make sure I understand what's happening.

Thanks, Bruce

@MrBruceKA, not a teacher exactly... at least not a school teacher like you might think.

Glad I could help. If you think my reply solves your issue, make sure to vote it up and mark it as the solution (if that is enabled for this post). That will flag the forum so others searching for it will be able to tell if your question was solved or not.

Let me know if you need anything else on FOLD. Like I said - it's tricky, but once you get it, it's pretty basic.