R help - Jan 2010 - Categorical data repeated on time analysis

Hi,

I am trying to analyze a data set when nematodes were killed after a
drug administration.

We have counted the number of nematode died and the number of nematode
survival at three time points.

So, there are 100% died in some plot and could be found zero percent
in another. Then, the data set have a lot of zeros.

I have googled and found a lot of information. Moreover, my data isn't
adjust to a normal distribution.

I have transformed it to square root, but, due to zeros, it don't fit
to a normal.

The design is a split-plot. I divided a Petri dish in four parts and
each day we measured one of they.

Here is a sample of the data:

trat	rep	time	killed	living	percent.killed	percent.living
1	1	48	8	6	57.14	42.86
1	2	48	17	15	53.13	46.88
1	3	48	6	4	60.00	40.00
1	1	72	17	15	53.13	46.88
1	2	72	24	33	42.11	57.89
1	3	72	11	0	100.00	0.00
1	1	96	18	28	39.13	60.87
1	2	96	19	6	76.00	24.00
1	3	96	9	10	47.37	52.63
2	1	48	7	2	77.78	22.22
2	2	48	10	4	71.43	28.57
2	3	48	8	2	80.00	20.00
2	1	72	5	2	71.43	28.57
2	2	72	14	13	51.85	48.15
2	3	72	30	1	96.77	3.23
2	1	96	2	6	25.00	75.00
2	2	96	11	15	42.31	57.69
2	3	96	3	2	60.00	40.00
3	1	48	8	8	50.00	50.00
3	2	48	6	7	46.15	53.85
3	3	48	0	2	0.00	100.00
3	1	72	3	3	50.00	50.00
3	2	72	5	1	83.33	16.67
3	3	72	18	10	64.29	35.71
3	1	96	4	0	100.00	0.00
3	2	96	0	0	0.00	0.00
3	3	96	18	19	48.65	51.35

We have counted killed and living because free-living nematode
reproduce very fast, so I need to know the number of living in the
medium.

What you suggest me for analyze this on R? What transformation I could do?

There were a specific package for that?

Have you did something like this?

Thank you very much

-- 
Marcelo Luiz de Laia
Lages - SC - Brazil
Linux user number 487797

Hi:

You might want to use a binomial generalized linear mixed model, for which
the
lme4 package and the functions lmer() and glmer() could come in handy. Since
there
is a dedicated list to mixed models, future questions on this particular
topic should
be sent there (subscribe to sig-mixed-models).

HTH,
Dennis

On Sun, Jan 24, 2010 at 2:26 PM, Marcelo Laia <marcelolaia@gmail.com>
wrote:
> Hi,
>
> I am trying to analyze a data set when nematodes were killed after a
> drug administration.
>
> We have counted the number of nematode died and the number of nematode
> survival at three time points.
>
> So, there are 100% died in some plot and could be found zero percent
> in another. Then, the data set have a lot of zeros.
>
> I have googled and found a lot of information. Moreover, my data isn't
> adjust to a normal distribution.
>
> I have transformed it to square root, but, due to zeros, it don't fit
> to a normal.
>
> The design is a split-plot. I divided a Petri dish in four parts and
> each day we measured one of they.
>
> Here is a sample of the data:
>
> trat    rep     time    killed  living  percent.killed  percent.living
> 1       1       48      8       6       57.14   42.86
> 1       2       48      17      15      53.13   46.88
> 1       3       48      6       4       60.00   40.00
> 1       1       72      17      15      53.13   46.88
> 1       2       72      24      33      42.11   57.89
> 1       3       72      11      0       100.00  0.00
> 1       1       96      18      28      39.13   60.87
> 1       2       96      19      6       76.00   24.00
> 1       3       96      9       10      47.37   52.63
> 2       1       48      7       2       77.78   22.22
> 2       2       48      10      4       71.43   28.57
> 2       3       48      8       2       80.00   20.00
> 2       1       72      5       2       71.43   28.57
> 2       2       72      14      13      51.85   48.15
> 2       3       72      30      1       96.77   3.23
> 2       1       96      2       6       25.00   75.00
> 2       2       96      11      15      42.31   57.69
> 2       3       96      3       2       60.00   40.00
> 3       1       48      8       8       50.00   50.00
> 3       2       48      6       7       46.15   53.85
> 3       3       48      0       2       0.00    100.00
> 3       1       72      3       3       50.00   50.00
> 3       2       72      5       1       83.33   16.67
> 3       3       72      18      10      64.29   35.71
> 3       1       96      4       0       100.00  0.00
> 3       2       96      0       0       0.00    0.00
> 3       3       96      18      19      48.65   51.35
>
> We have counted killed and living because free-living nematode
> reproduce very fast, so I need to know the number of living in the
> medium.
>
> What you suggest me for analyze this on R? What transformation I could do?
>
> There were a specific package for that?
>
> Have you did something like this?
>
> Thank you very much
>
> --
> Marcelo Luiz de Laia
> Lages - SC - Brazil
> Linux user number 487797
>
> ______________________________________________
> R-help@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
	[[alternative HTML version deleted]]