@nama 

The possibility of equality is always part of the null hypothesis and never part of the alternative hypothesis.

@nama 

The command

Statistics:-TwoSamplePairedTTest(X, Z, 0, confidence= 0.95, alternative= 'lowertail')

means that the alternative hypothesis is that the first factor (X) is less than the second factor (Z).

What input did you call the procedure with?

Please either post the procedure as plaintext so that I can cut and paste it or attach it as a worksheet so that I can download it.

Do you have initial values for lambda1, lambda2, and lambda3?

What exactly is the problem that you're having? When I run your code, there is no apparent problem.

@lazza For any entry like 1+10^(-10), first apply evalf to the matrix, then fnormal:

map(fnormal, evalf(M));

Your file did not attach correctly; the link is dead. Please try uploading again.

@mahmood180 It is true that diff(f(t), t$0) is undefined; however ThU said D[1$m](f)(t). That is defined, correctly, at m=0.

This is not related to your question, but you should spell Pi with an uppercase P in Maple. You will get a much simpler answer, to wit,

as opposed to the three-line answer that you had before. With a lowercase p, pi is just an arbitrary name with no predefined meaning.

@mahmood180 

Preben said

g1:= unapply(...);

but your code has

g1(t):= unapply(...);

As Acer said, you should write the coefficients directly to a Matrix and skip entirely the equation step. Extracting the coefficients from large equations is going to be inefficient even in the best case.

@Carl Love I'll need to see that in a worksheet to diagnose it. My first guess is that you mistakenly used square brackets in a place where you should've used parentheses.

@lt In the lines where you define the numeric values, you did not use := for g and h.

@lt Yes, you are correct on both points.

In all cases, the result is returned as a list.

@fairuzalwani 

1 +~ map2(ListTools:-BinaryPlace, E, [3, 10, 24]);