f309a043-4f30-43bc-aaf7-8c9d926f0441 7 4 5 decimal

&[DATE] &[TIME] - &[FILENAME]

The "node e" model function n x f x n^x exp 1 - / :

ε 1015 -^:

plot 0.75 x f 1 x f 2 x f 2.5 x f 3 x f 0.0625 5 x f (* 1.5 x f 71 sys :

1. 0.581976706869326 =>1/(e-1) ... a constant in this model 2. [1,0.582,"+",20,"black"] has virtue of constraining. 3. impairs: divide by 0, can not calculate ... rejects X_Y

plot 1 0.582 + 20 black 1 5 mat 2 1 sys

0.75 1 f 11 f 21 f 2.5 1 f 31 f 51 f 61 sys 0.581976706869326 0.581976706869326 0.581976706869326 0.581976706869326 0.581976706869326 0.581976706869326 61 sys

2.5 x f x ε 10 int 3.726

Σcumul cumul integration x ε 10 ε 0.1 + (range : i 1 x rows range Σ i el 2 t f t x 1 el x i el int : for x Σ augment 41 line :

Σcumul

integrate wisely low region 0.75 x f x 0.5 10 int 1.0909

Read this about PDF:

If you can model your data set by some of the f(n>1,x) ... it is tempting to figure a disribution,

i.e: a personalised distribution. This is possible ... you would have to figure the Σcumul @ ∞

then scale to unity. As exemplified, none is a PDF ... to be a PDF, the Σcumul must be

asymptotic to 1 @ ∞.

]]>