I am trying to use the package “unmarked” under R, and especially the distsamp() function following the paper written by Richard Chandler (January 13, 2014). I have no problem having the data imported, formatted and summarized, and my histogram of detection distances is ok. However, when I want to fit the model: hn_Null<-distsamp(~1~1,umf) I get the error: "Hessian is singular. Try using fewer covariates or providing starting values." I can’t figure why! I have 7 transect, around 60 species and more than 3000 observations. I have no transect-specific covariates. My data has the form:
|
Transect |
Length |
Species |
Number |
Distance |
|
a |
19900 |
Grand Cormoran |
1 |
163 |
|
a |
19900 |
Fou de Bassan |
1 |
272 |
|
a |
19900 |
Fou de Bassan |
1 |
131 |
|
a |
19900 |
Fou de Bassan |
1 |
218 |
|
a |
19900 |
Mouette mélanocéphale |
1 |
54 |
|
a |
19900 |
Fou de Bassan |
1 |
1088 |
And my umf object is:
Data frame representation of unmarkedFrame object.
y.1 y.2 y.3 y.4 y.5 y.6
a 273 163 61 101 59 57
b 42 16 14 8 2 6
c 298 150 96 107 39 74
d 82 42 20 26 15 19
e 293 153 91 89 49 61
f 46 21 14 22 8 11
g 202 120 68 76 35 58
I have tried to supply starting values but with no result.
Do you have any idea/suggestion to solve that problem?
Thank you in advance.
Best regards.