distsamp() function

117 views
Skip to first unread message

Hervé Darzacq

unread,
Apr 18, 2014, 8:10:44 AM4/18/14
to unma...@googlegroups.com
Hi all,

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.

Alex Anderson

unread,
Apr 30, 2014, 7:05:54 AM4/30/14
to unma...@googlegroups.com
Hi all,
I am currently working on some nocturnal forest mammal data (from
spotlight surveys) using pcountOPEN. I have 35 primary periods (defined
by year/season combinations) as that was the period in which I felt it
reasonable to assume closure. However, I am really interested in
differences between two longer time-periods distinguished before and
after primary period 16 or so. ("time_step3" in my model). This may be
a simple question, but is there are recommended way to summarize such
population trends at a higher level than primary period perhaps using
the yearlySiteCovs? an example model looks like this:

(fm37a <- pcountOpen(~MATemp+MATemp2+annual_pptn12, ~time_step3,
~time_step3,
~wet+wind+season+temp_anomaly+moon+decimal_start, spp.umf,
K=tk,mixture = "ZIP", dynamics = "constant", se =
TRUE))

and an unmarked frame looks like this:

137 sites
Maximum number of observations per site: 280
Mean number of observations per site: 5.69
Number of primary survey periods: 35
Number of secondary survey periods: 8
Sites with at least one detection: 26

Tabulation of y observations:
0 1 2 3 4 5 6 7 8 9 10 11 12
565 21 20 21 12 17 20 11 11 11 10 5 9
13 14 15 16 17 19 20 21 22 23 24 25 26
7 3 5 4 4 1 6 2 1 2 3 1 2
27 28 34 <NA>
3 1 2 37580

Site-level covariates:
point_id MATemp.V1 accu04.V1 annual_pptn12.V1
AU10A0 : 1 Min. :-1.8429006 Min. :-3.416402 Min. :-1.313057
AU10B0 : 1 1st Qu.:-0.8695779 1st Qu.:-0.810225 1st Qu.:-0.728798
AU1A1 : 1 Median : 0.1418747 Median : 0.275329 Median :-0.183983
AU2A0 : 1 Mean : 0.0000000 Mean : 0.000000 Mean : 0.000000
AU4A0 : 1 3rd Qu.: 0.9061677 3rd Qu.: 0.797857 3rd Qu.: 0.712126
AU4B0 : 1 Max. : 1.7216363 Max. : 1.669537 Max. : 4.753864
(Other):131
latdd.V1 slope.V1 aspect.V1
Min. :-1.6176393 Min. :-1.006430 Min. :-1.6186105
1st Qu.:-0.9236487 1st Qu.:-0.713359 1st Qu.:-0.7883060
Median :-0.1513666 Median :-0.273280 Median : 0.1463615
Mean : 0.0000000 Mean : 0.000000 Mean : 0.0000000
3rd Qu.: 0.9638133 3rd Qu.: 0.447349 3rd Qu.: 0.8260523
Max. : 1.8799902 Max. : 3.923585 Max. : 1.5326372

dem_250.V1 time_step3
Min. :-2.4119946 Min. :1.000
1st Qu.:-0.5562269 1st Qu.:1.000
Median : 0.0127130 Median :1.000
Mean : 0.0000000 Mean :1.264
3rd Qu.: 0.6530920 3rd Qu.:1.500
Max. : 1.9696326 Max. :2.000


Observation-level covariates:
X point_id occasion2 point_primary
Min. : 1 AU10A0 : 280 Min. : 1.100 CU12A0_26: 8
1st Qu.: 9591 AU10B0 : 280 1st Qu.: 9.675 AUSE3_4 : 7
Median :19180 AU1A1 : 280 Median :18.450 CU12B0_26: 7
Mean :19180 AU2A0 : 280 Mean :18.450 AU10A0_2 : 6
3rd Qu.:28770 AU4A0 : 280 3rd Qu.:27.225 AU10A0_4 : 6
Max. :38360 AU4B0 : 280 Max. :35.800 (Other) : 746
(Other):36680 NA's :37580
occasion wind wet.V1 mist
Min. :1.00 Min. :0.00 Min. :-1.49 Min. :0.00
1st Qu.:1.00 1st Qu.:0.00 1st Qu.:-0.49 1st Qu.:0.00
Median :1.00 Median :1.00 Median :-0.49 Median :0.00
Mean :1.77 Mean :1.08 Mean : 0.00 Mean :0.28
3rd Qu.:2.00 3rd Qu.:1.00 3rd Qu.: 0.52 3rd Qu.:0.00
Max. :8.00 Max. :6.00 Max. : 2.54 Max. :3.00
NA's :37580 NA's :37580 NA's :37580 NA's :37580
temp_anomaly.V1 moon month decimal_start
Min. :-3.69 Min. :0.00 Min. : 1.00 Min. :18.50
1st Qu.:-0.48 1st Qu.:0.00 1st Qu.: 3.00 1st Qu.:19.75
Median : 0.03 Median :0.00 Median : 6.00 Median :20.04
Mean : 0.00 Mean :0.23 Mean : 6.09 Mean :20.06
3rd Qu.: 0.56 3rd Qu.:0.00 3rd Qu.:10.00 3rd Qu.:20.25
Max. : 2.83 Max. :2.00 Max. :12.00 Max. :23.67
NA's :37580 NA's :37580 NA's :37580 NA's :37580
season time_step3 year julian_date
Min. :1.00 Min. :1.00 Min. :1992 Min. : 828
1st Qu.:1.00 1st Qu.:1.00 1st Qu.:1999 1st Qu.:3215
Median :2.00 Median :2.00 Median :2006 Median :5892
Mean :1.57 Mean :1.51 Mean :2004 Mean :5227
3rd Qu.:2.00 3rd Qu.:2.00 3rd Qu.:2010 3rd Qu.:7281
Max. :2.00 Max. :2.00 Max. :2013 Max. :8624
NA's :37580 NA's :37580 NA's :37580 NA's :37580

Yearly-site-level covariates:
point_id primary_period point_primary time_step3
AU10A0 : 35 Min. : 1 AU10A0_1 : 1 Min. :1.000
AU10B0 : 35 1st Qu.: 9 AU10A0_10: 1 1st Qu.:1.000
AU1A1 : 35 Median :18 AU10A0_11: 1 Median :1.000
AU2A0 : 35 Mean :18 AU10A0_12: 1 Mean :1.429
AU4A0 : 35 3rd Qu.:27 AU10A0_13: 1 3rd Qu.:2.000
AU4B0 : 35 Max. :35 AU10A0_14: 1 Max. :2.000
(Other):4585 (Other) :4789

Any thoughts would be much appreciated.
regards,
Alex
Reply all
Reply to author
Forward
0 new messages