Not a Number (NaN) for the standard error, lower and upper using predict function in unmarked

128 views Asked by At

Why do I get NaNs when running the following code? I am conducting a distance sampling analysis of species density per covariate. This hasn't happened for any of my other covariates?

#summary of umf2

> summary(umf2)
unmarkedFrameDS Object

line-transect survey design
Distance class cutpoints (m):  0 5 10 15 20 25 30 35 

64 sites
Maximum number of distance classes per site: 7 
Mean number of distance classes per site: 7 
Sites with at least one detection: 8 

Tabulation of y observations:
  0   1   2 
434  12   2 

Site-level covariates:
           transect     length_m        length          number     
 01_Cungha_T1  : 1   Min.   : 996   Min.   :0.996   Min.   : 1.00  
 02_Capicada_T1: 1   1st Qu.:1061   1st Qu.:1.061   1st Qu.:16.75  
 03_Caghode_T1 : 1   Median :1098   Median :1.099   Median :32.50  
 04_Caghode_T2 : 1   Mean   :1126   Mean   :1.126   Mean   :32.50  
 05_Cafal_T1   : 1   3rd Qu.:1167   3rd Qu.:1.167   3rd Qu.:48.25  
 06_Muna_T1    : 1   Max.   :1758   Max.   :1.758   Max.   :64.00  
 (Other)       :58                                                  
                                                                    
       y               x            d_forest_1    d_all_roads_1   
 Min.   :11.10   Min.   :-15.16   Min.   :    0   Min.   :   5.0  
 1st Qu.:11.19   1st Qu.:-15.08   1st Qu.:    0   1st Qu.: 180.3  
 Median :11.25   Median :-15.02   Median :   10   Median : 534.8  
 Mean   :11.27   Mean   :-15.00   Mean   : 2751   Mean   : 811.5  
 3rd Qu.:11.34   3rd Qu.:-14.94   3rd Qu.: 2893   3rd Qu.:1145.9  
 Max.   :11.42   Max.   :-14.77   Max.   :17242   Max.   :3666.7    
                                                               

#fitting an a priori model set: all roads, hazard.

m.haz.2.allroads <- distsamp(~1 ~d_all_roads_1, umf2, keyfun="hazard", output="density", unitsOut="kmsq")

#predict with distance to all roads

> m.allroads2 <-  data.frame(d_all_roads_1=seq(5.0000, 3666.6775, length=64))    
> allroads.pred2 <- predict(m.haz.2.allroads, type="state", newdata=m.allroads2, appendData=TRUE)
There were 50 or more warnings (use warnings() to see the first 50)
> allroads.pred2 

   Predicted         SE    lower    upper d_all_roads_1
1   1.979158 0.23962464 1.561069 2.509221       5.00000
2   2.041778 0.23667781 1.626820 2.562582      63.12187
3   2.106379 0.23300058 1.695819 2.616337     121.24373
4   2.173025 0.22849261 1.768321 2.670350     179.36560
5   2.241779 0.22303220 1.844623 2.724444     237.48746
...
10  2.619589 0.17424708 2.299396 2.984369     528.09679
11  2.702472 0.15742845 2.410881 3.029331     586.21865
12  2.787978 0.13611667 2.533561 3.067943     644.34052
13  2.876189 0.10742754 2.673157 3.094641     702.46238
14  2.967191 0.06136485 2.849323 3.089934     760.58425
15  3.061072        NaN      NaN      NaN     818.70611
16  3.157924        NaN      NaN      NaN     876.82798
17  3.257839        NaN      NaN      NaN     934.94984
18  3.360917        NaN      NaN      NaN     993.07171
19  3.467255        NaN      NaN      NaN    1051.19357
...
60 12.434748        NaN      NaN      NaN    3434.19004
61 12.828180        NaN      NaN      NaN    3492.31190
62 13.234061        NaN      NaN      NaN    3550.43377
63 13.652784        NaN      NaN      NaN    3608.55563
64 14.084755        NaN      NaN      NaN    3666.67750

Please let me know if any further information is needed to help me solve this, many thanks.

0

There are 0 answers