Linear Regression with Tradestation

May 4th, 2019, 04:37 PM

Hello All - I have looked through a few threads but have not found anything that helps in my understanding of the current problem I am running into. I use the tradestation platform. I am trying to understand how the linear regression function works.

To create my first test, I am attempting to regress the last 30 bars of SP 500 Emini (dependent; data1 feed) versus the Russel (independent; data4 feed). I used the resources on the tradestation website to help construct the code below. However, when looking at the output variables, the numbers do not make sense. I pulled them all into a print file and calculated the regression manually in excel. I tried to troubleshoot but cannot figure out what I am doing wrong.

Any suggestions are appreciated.

Thank you!

----------------Start Code: My Failed Attempt------------------
Inputs:
RegArrayLength(30);

Var:
x(0), oLRSlope(0), oLRAngle(0), oLRIntercept(0), oLRValueRaw(0);

Array: ES_PrevCloses[30](0), RUS_PrevCloses[30](0);

For x = 1 to RegArrayLength begin

ES_PrevCloses[x] = Close[x] of data1;
RUS_PrevCloses[x] = Close[x] of data4;

End;

Value1 = LinRegArray2(RUS_PrevCloses, ES_PrevCloses, RegArrayLength, 0, oLRSlope, oLRAngle, oLRIntercept, oLRValueRaw);

Value2 = oLRSlope;

Value3 = oLRAngle;

Value4 = oLRIntercept;

Value5 = oLRValueRaw;

Value6 = Value4 + ((close[1] of data4)* Value2);

---------------------END Code---------------------

May 7th, 2019, 07:30 AM

I don't know easy language but part of the problem is easy language is just outdated. Python is an easier language than easy language.
If you have your prices in X and Y

from sklearn.linear_model import LinearRegression
ln = LinearRegression()
ln.fit(X,Y)

I just don't think it is worth the time to learn easy language in 2019.

Linear Regression with Tradestation

Discussion in Traders Hideout

Linear Regression with Tradestation