Simple Fibonacci calculation

re: Simple Fibonacci calculation

Paul Nielsen — Sun, 18 Oct 2009 14:16:30 GMT

very cool. well done.

-Pn

re: Simple Fibonacci calculation

John Meriweather — Mon, 19 Oct 2009 17:53:16 GMT

What's the point of calculating the Fibonacci series in T-SQL? And why would anyone use t-SQL for that calculation?

re: Simple Fibonacci calculation

Jason — Tue, 20 Oct 2009 15:00:44 GMT

@Peter: This is a great example of a bad post.

First of all, you only provide a heap of code with no explanation as to why it is interesting (thanks Peso) or how you might use it.

Second, your code essentially implements the addition operator for strings instead of just using the built in "+", which is just stupid.

@Paul: This post is neither cool nor well done.

re: Simple Fibonacci calculation

Linchi Shea — Tue, 20 Oct 2009 17:25:16 GMT

I don't think John was questioning the significance of the Fibonacci series, but rather why one would do it with T-SQL. For classroom exercises and seminar demos, it's fine. It's also fine if one just wants to kill some time. But I'd agree that people should probably be cautioned not to actually end up writing production T-SQL code to do Fibonacci searies. Well, maybe there is a real case where writing it in T-SQL is justified. But it is certianly difficult to come up with one.

re: Simple Fibonacci calculation

Peso — Tue, 20 Oct 2009 18:54:00 GMT

New content posted.

re: Simple Fibonacci calculation

Jason — Wed, 21 Oct 2009 13:51:18 GMT

Hi Peter,

I like the revisions you made to this article. Much more interesting.

As an alternative to BIGINT, you could use DECIMAL(38,0) and calculate up to the 184th number in the sequence.

Thanks, Jason

re: Simple Fibonacci calculation

Joe Celko — Sun, 25 Oct 2009 13:49:50 GMT

Wouldn't it be faster to use the constant phi and the closed form?

Fibonacci(@n INTEGER)

RETURN

EOUND (((POWER (1.6190339887, @n)- POWER (1.0 - 1.6190339887, @n))/

SQRT (5.0)), 0);

untested.

re: Simple Fibonacci calculation

Peso — Sun, 25 Oct 2009 16:28:46 GMT

It would be great if it worked! However, the golden ration is the ration for which division of Fibonacci number converges to, when sufficiently large.

However, the formula you displayed above gives wrong answer very early, and after step 12, the error is larger and larger

n Peso Celko

-- ---- -----

1 0 0

2 1 1

3 1 1

4 2 2

5 3 3

6 5 5

7 8 8

8 13 13

9 21 21

10 34 34

11 55 55

12 89 90

13 144 145

re: Simple Fibonacci calculation

RBArryYoung — Mon, 26 Oct 2009 06:27:40 GMT

This one is accurate up to n=62:

FLOOR(POWER(1.61803398874989, n)/SQRT(5) +.5)

:-)

re: Simple Fibonacci calculation

Peso — Mon, 26 Oct 2009 06:52:24 GMT

Much better.