-
Notifications
You must be signed in to change notification settings - Fork 0
/
todo.txt
57 lines (45 loc) · 2.36 KB
/
todo.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
package/function names
- deft (detection efficiency for telemetry)
Readme/articles
- reference to gold standard range testing methods (and why pkg works for that)
- move detailed readme sections into articles for website (only include simple call to dr_fit and installation)
- articles
- getting started with detrange (includes background)
- doing more with detrange
- working with dr_template - changing priors, param names etc.
- working with samples/model objects
- analysis coping with CPDI (i.e. quadratic relationship)
- simulate hypothetical DR or known DR reported in other studies (i.e. no data all priors)
Issues/fixes
- get edu/joe to look over term descriptions
- ensure no negative distance estimates - how to properly bound? including lower/upper est
Enhancements
- calculate corrected distance from receiver/tag depth
- function to plot posteriors of distance estimates?
- build shiny app
- put on CRAN
- link to funs in documentation
- include env covariates
- fully document datasets
DR_FIT
MODEL
- implement correlated slope/intercept model
- model for nested station/transect (i.e. cardinal directions at single station)
- model for CPDI - lower detection efficiency at close distance due to CPDI (i.e. a n-shaped curve), should fit two models (one with breakpoint) and use AIC to select the best model - or model average for estimates.
- covariates - term names = column name prefix with b + s<>Station + b<>Station
- add function arg to control # iterations, e.g. c('quick' (500), 'report' (1000), 'paper' (1500))
- standardize covariate?
GENERICS
- Use generics for model object manipulation - some ideas:
- explore (run shiny app/widget on object) invokes a function that starts an interactive, pre-defined widget (e.g. plotly visualization, shiny app, etc.) to investigate the results. - could be used to open leaflet map of sites and detection ranges (alpha by de + buffer by distance)
- visualize (maybe plot? or use autoplot?) - Visualize a data set or object.
- subset (if fit multiple models)
done: augment, tidy, glance, predict, summary, estimates, coef, autoplot, nchains, npars, nterms, niters, rhat, esr, converged, pars
PLOTS
fix argument names of geom_dr_errorbar (lower, estimate, upper)
add trace and density plot functions
Pkg maintenance stuff
- build website
- set up GH actions and badge
- add lifecycle badges
- add test coverage