R help - Oct 2014 - glm - rgamma vector as predictor

I have a strange question concerning the fit of a Gamma generalized linear
model with glm (and further using gamma.shape to measure the shape
parameter).

Actually, I started with rgamma to generate some random vectors because I
wanted to play around with various conditions (and become familiar with
gamma shape). But before you can start with gamma shape you need to have a
glm object.

Assuming

gamma.random <- rgamma (1000,1.5)

is my random vector.

How do I create a glm object if I only have one vector? I guess, my random
vector is the predictor and the response is the probability (in the sense
of a pdf). Can anybody give me a hint how the syntax is?

glm (? ~ gamma.random, family=Gamma)

Thanks Wim

	[[alternative HTML version deleted]]

Wim Kreinen <wkreinen <at> gmail.com> writes:
>
[snip]
> 
> Actually, I started with rgamma to generate some random vectors because I
> wanted to play around with various conditions (and become familiar with
> gamma shape). But before you can start with gamma shape you need to have a
> glm object.
> 
> Assuming
> 
> gamma.random <- rgamma (1000,1.5)
> 
> is my random vector.
> 
> How do I create a glm object if I only have one vector? I guess, my random
> vector is the predictor and the response is the probability (in the sense
> of a pdf). Can anybody give me a hint how the syntax is?
 [snip]

I think you're looking for

  MASS::gamma.shape(glm(gamma.random ~ 1, family=Gamma))