Mercurial > repos > ecology > regionalgam_flight_curve
comparison dennis-gam-initial-functions.R @ 0:fff507f00b2b draft default tip
planemo upload for repository https://github.com/galaxyecology/tools-ecology/tree/master/tools/regionalgam commit ffe42225fff8992501b743ebe2c78e50fddc4a4e
author | ecology |
---|---|
date | Thu, 20 Jun 2019 04:03:07 -0400 |
parents | |
children |
comparison
equal
deleted
inserted
replaced
-1:000000000000 | 0:fff507f00b2b |
---|---|
1 ### R-Script Adapted from script provided by the CEH, UK BY: Reto Schmucki [ reto.schmucki@mail.mcgill.ca] | |
2 ### DATE: 14 July 2014 function to run two stage model in DENNIS et al. 2013 | |
3 | |
4 | |
5 .onAttach <- function(libname, pkgname) { | |
6 packageStartupMessage(" The regionalGAM package that is no longer maintained, \n use the new rbms package instead. \n | |
7 devtools::install_github(\"RetoSchmucki/rbms\", force=TRUE)") | |
8 } | |
9 | |
10 | |
11 #' year_day_func Function | |
12 #' This function generate the full sequence of days, months and include the observation to that file. | |
13 #' @param sp_data A data.frame with your observation. | |
14 #' @keywords year days | |
15 #' @export | |
16 #' @author Reto Schmucki | |
17 #' @examples | |
18 #' year_day_func() | |
19 | |
20 | |
21 # FUNCTIONS | |
22 | |
23 year_day_func = function(sp_data) { | |
24 | |
25 year <- unique(sp_data$YEAR) | |
26 | |
27 origin.d <- paste(year, "01-01", sep = "-") | |
28 if ((year%%4 == 0) & ((year%%100 != 0) | (year%%400 == 0))) { | |
29 nday <- 366 | |
30 } else { | |
31 nday <- 365 | |
32 } | |
33 | |
34 date.serie <- as.POSIXlt(seq(as.Date(origin.d), length = nday, by = "day"), format = "%Y-%m-%d") | |
35 | |
36 dayno <- as.numeric(julian(date.serie, origin = as.Date(origin.d)) + 1) | |
37 month <- as.numeric(strftime(date.serie, format = "%m")) | |
38 week <- as.numeric(strftime(date.serie, format = "%W")) | |
39 week_day <- as.numeric(strftime(date.serie, format = "%u")) | |
40 day <- as.numeric(strftime(date.serie, format = "%d")) | |
41 | |
42 site_list <- sp_data[!duplicated(sp_data$SITE), c("SITE")] | |
43 | |
44 all_day_site <- data.frame(SPECIES = sp_data$SPECIES[1], SITE = rep(site_list, rep(nday, length(site_list))), | |
45 YEAR = sp_data$YEAR[1], MONTH = month, WEEK = week, DAY = day, DAY_WEEK = week_day, DAYNO = dayno, | |
46 COUNT = NA) | |
47 | |
48 count_index <- match(paste(sp_data$SITE, sp_data$DAYNO, sep = "_"), paste(all_day_site$SITE, all_day_site$DAYNO, | |
49 sep = "_")) | |
50 all_day_site$COUNT[count_index] <- sp_data$COUNT | |
51 site_count_length <- aggregate(sp_data$COUNT, by = list(sp_data$SITE), function(x) list(1:length(x))) | |
52 names(site_count_length$x) <- as.character(site_count_length$Group.1) | |
53 site_countno <- utils::stack(site_count_length$x) | |
54 all_day_site$COUNTNO <- NA | |
55 all_day_site$COUNTNO[count_index] <- site_countno$values # add count number to ease extraction of single count | |
56 | |
57 # Add zero to close observation season two weeks before and after the first and last | |
58 first_obs <- min(all_day_site$DAYNO[!is.na(all_day_site$COUNT)]) | |
59 last_obs <- max(all_day_site$DAYNO[!is.na(all_day_site$COUNT)]) | |
60 | |
61 closing_season <- c((first_obs - 11):(first_obs - 7), (last_obs + 7):(last_obs + 11)) | |
62 | |
63 # If closing season is before day 1 or day 365, simply set the first and last 5 days to 0 | |
64 if (min(closing_season) < 1) | |
65 closing_season[1:5] <- c(1:5) | |
66 if (max(closing_season) > nday) | |
67 closing_season[6:10] <- c((nday - 4):nday) | |
68 | |
69 all_day_site$COUNT[all_day_site$DAYNO %in% closing_season] <- 0 | |
70 all_day_site$ANCHOR <- 0 | |
71 all_day_site$ANCHOR[all_day_site$DAYNO %in% closing_season] <- 1 | |
72 | |
73 all_day_site <- all_day_site[order(all_day_site$SITE, all_day_site$DAYNO), ] | |
74 | |
75 return(all_day_site) | |
76 } | |
77 | |
78 | |
79 #' trap_area Function | |
80 #' | |
81 #' This function compute the area under the curve using the trapezoid method. | |
82 #' @param x A vector or a two columns matrix. | |
83 #' @param y A vector, Default is NULL | |
84 #' @keywords trapezoid | |
85 #' @export | |
86 #' @examples | |
87 #' trap_area() | |
88 | |
89 | |
90 trap_area = function(x, y = NULL) { | |
91 # If y is null and x has multiple columns then set y to x[,2] and x to x[,1] | |
92 if (is.null(y)) { | |
93 if (length(dim(x)) == 2) { | |
94 y = x[, 2] | |
95 x = x[, 1] | |
96 } else { | |
97 stop("ERROR: need to either specifiy both x and y or supply a two column data.frame/matrix to x") | |
98 } | |
99 } | |
100 | |
101 # Check x and y are same length | |
102 if (length(x) != length(y)) { | |
103 stop("ERROR: x and y need to be the same length") | |
104 } | |
105 | |
106 # Need to exclude any pairs that are NA for either x or y | |
107 rm_inds = which(is.na(x) | is.na(y)) | |
108 if (length(rm_inds) > 0) { | |
109 x = x[-rm_inds] | |
110 y = y[-rm_inds] | |
111 } | |
112 | |
113 # Determine values of trapezoids under curve Get inds | |
114 inds = 1:(length(x) - 1) | |
115 # Determine area using trapezoidal rule Area = ( (b1 + b2)/2 ) * h where b1 and b2 are lengths of bases | |
116 # (the parallel sides) and h is the height (the perpendicular distance between two bases) | |
117 areas = ((y[inds] + y[inds + 1])/2) * diff(x) | |
118 | |
119 # total area is sum of all trapezoid areas | |
120 tot_area = sum(areas) | |
121 | |
122 # Return total area | |
123 return(tot_area) | |
124 } | |
125 | |
126 | |
127 #' trap_index Function | |
128 #' | |
129 #' This function compute the area under the curve (Abundance Index) across species, sites and years | |
130 #' @param sp_data A data.frame containing species count data generated from the year_day_func() | |
131 #' @param y A vector, Default is NULL | |
132 #' @keywords Abundance index | |
133 #' @export | |
134 #' @examples | |
135 #' trap_index() | |
136 | |
137 | |
138 | |
139 trap_index = function(sp_data, data_col = "IMP", time_col = "DAYNO", by_col = c("SPECIES", "SITE", "YEAR")) { | |
140 | |
141 # Build output data.frame | |
142 out_obj = unique(sp_data[, by_col]) | |
143 # Set row.names to be equal to collapsing of output rows (will be unique, you need them to make uploading | |
144 # values back to data.frame will be easier) | |
145 row.names(out_obj) = apply(out_obj, 1, paste, collapse = "_") | |
146 | |
147 # Using this row.names from out_obj above as index in by function to loop through values all unique combs | |
148 # of by_cols and fit trap_area to data | |
149 ind_dat = by(sp_data[, c(time_col, data_col)], apply(sp_data[, by_col], 1, paste, collapse = "_"), trap_area) | |
150 | |
151 # Add this data to output object | |
152 out_obj[names(ind_dat), "SINDEX"] = round(ind_dat/7, 1) | |
153 | |
154 # Set row.names to defaults | |
155 row.names(out_obj) = NULL | |
156 | |
157 # Return output object | |
158 return(out_obj) | |
159 } | |
160 | |
161 | |
162 #' flight_curve Function | |
163 #' This function compute the flight curve across and years | |
164 #' @param your_dataset A data.frame containing original species count data. The data format is a csv (comma "," separated) with 6 columns with headers, namely "species","transect_id","visit_year","visit_month","visit_day","sp_count" | |
165 #' @keywords standardize flight curve | |
166 #' @export | |
167 #' @examples | |
168 #' flight_curve() | |
169 | |
170 | |
171 flight_curve <- function(your_dataset, GamFamily = 'nb', MinVisit = 2, MinOccur = 1) { | |
172 | |
173 if("mgcv" %in% installed.packages() == "FALSE") { | |
174 print("mgcv package is not installed.") | |
175 x <- readline("Do you want to install it? Y/N") | |
176 if (x == 'Y') { | |
177 install.packages("mgcv") | |
178 } | |
179 if (x == 'N') { | |
180 stop("flight curve can not be computed without the mgcv package, sorry") | |
181 } | |
182 } | |
183 | |
184 flight_pheno <- data.frame() | |
185 | |
186 your_dataset$DAYNO <- strptime(paste(your_dataset$DAY, your_dataset$MONTH, | |
187 your_dataset$YEAR, sep = "/"), "%d/%m/%Y")$yday + 1 | |
188 dataset <- your_dataset[, c("SPECIES", "SITE", "YEAR", "MONTH", | |
189 "DAY", "DAYNO", "COUNT")] | |
190 sample_year <- unique(dataset$YEAR) | |
191 sample_year <- sample_year[order(sample_year)] | |
192 | |
193 if (length(sample_year) >1 ) { | |
194 for (y in sample_year) { | |
195 dataset_y <- dataset[dataset$YEAR == y, ] | |
196 | |
197 # subset sites with enough visit and occurence | |
198 occ <- aggregate(dataset_y$COUNT, by = list(SITE = dataset_y$SITE), function(x) sum(x > 0)) | |
199 vis <- aggregate(dataset_y$COUNT, by = list(SITE = dataset_y$SITE), function(x) length(x)) | |
200 dataset_y <- dataset_y[dataset_y$SITE %in% occ$SITE[occ$x >= MinOccur], ] | |
201 dataset_y <- dataset_y[dataset_y$SITE %in% vis$SITE[vis$x >= MinVisit], ] | |
202 nsite <- length(unique(dataset_y$SITE)) | |
203 if (nsite < 1) { | |
204 print(paste("No sites with sufficient visits and occurence, MinOccur:", MinOccur, " MinVisit: ", MinVisit, " for " , dataset$SPECIES[1],"at year", y)) | |
205 next | |
206 } | |
207 # Determine missing days and add to dataset | |
208 sp_data_all <- year_day_func(dataset_y) | |
209 if (nsite > 200) { | |
210 sp_data_all <- sp_data_all[as.character(sp_data_all$SITE) %in% as.character(unique(dataset_y$SITE)[sample(1:nsite, | |
211 200, replace = F)]), ] | |
212 sp_data_all <- sp_data_all | |
213 } | |
214 sp_data_all$trimDAYNO <- sp_data_all$DAYNO - min(sp_data_all$DAYNO) + 1 | |
215 print(paste("Fitting the GAM for",as.character(sp_data_all$SPECIES[1]),"and year",y,"with",length(unique(sp_data_all$SITE)),"sites :",Sys.time())) | |
216 if(length(unique(sp_data_all$SITE))>1){ | |
217 gam_obj_site <- try(mgcv::gam(COUNT ~ s(trimDAYNO, bs = "cr") + as.factor(SITE) -1, | |
218 data = sp_data_all, family = GamFamily), silent = TRUE) | |
219 } | |
220 else { | |
221 gam_obj_site <- try(mgcv::gam(COUNT ~ s(trimDAYNO, bs = "cr") -1, | |
222 data = sp_data_all, family = GamFamily), silent = TRUE) | |
223 } | |
224 # Give a second try if the GAM does not converge the first time | |
225 if (class(gam_obj_site)[1] == "try-error") { | |
226 # Determine missing days and add to dataset | |
227 sp_data_all <- year_day_func(dataset_y) | |
228 if (nsite > 200) { | |
229 sp_data_all <- sp_data_all[as.character(sp_data_all$SITE) %in% as.character(unique(dataset_y$SITE)[sample(1:nsite, | |
230 200, replace = F)]), ] | |
231 } | |
232 else { | |
233 sp_data_all <- sp_data_all | |
234 } | |
235 sp_data_all$trimDAYNO <- sp_data_all$DAYNO - min(sp_data_all$DAYNO) + 1 | |
236 print(paste("Fitting the GAM for",sp_data_all$SPECIES[1],"at year", y,"with",length(unique(sp_data_all$SITE)),"sites :",Sys.time(),"second try")) | |
237 if(length(unique(sp_data_all$SITE))>1){ | |
238 gam_obj_site <- try(mgcv::gam(COUNT ~ s(trimDAYNO, bs = "cr") + as.factor(SITE) -1, | |
239 data = sp_data_all, family = GamFamily), silent = TRUE) | |
240 } | |
241 else { | |
242 gam_obj_site <- try(mgcv::gam(COUNT ~ s(trimDAYNO, bs = "cr") -1, | |
243 data = sp_data_all, family = GamFamily), silent = TRUE) | |
244 } | |
245 if (class(gam_obj_site)[1] == "try-error") { | |
246 print(paste("Error in fitting the flight period for",sp_data_all$SPECIES[1],"at year", y,"no convergence after two trial")) | |
247 sp_data_all[, "FITTED"] <- NA | |
248 sp_data_all[, "COUNT_IMPUTED"] <- NA | |
249 sp_data_all[is.na(sp_data_all$COUNT), "COUNT_IMPUTED"] <- NA | |
250 sp_data_all[, "NM"] <- NA | |
251 } | |
252 else { | |
253 # Generate a list of values for all days from the additive model and use | |
254 # these value to fill the missing observations | |
255 sp_data_all[, "FITTED"] <- mgcv::predict.gam(gam_obj_site, newdata = sp_data_all[, | |
256 c("trimDAYNO", "SITE")], type = "response") | |
257 # force zeros at the beginning end end of the year | |
258 sp_data_all[sp_data_all$trimDAYNO < 60, "FITTED"] <- 0 | |
259 sp_data_all[sp_data_all$trimDAYNO > 305, "FITTED"] <- 0 | |
260 # if infinite number in predict replace with NA. | |
261 if(sum(is.infinite(sp_data_all[, "FITTED"]))>0){ | |
262 print(paste("Error in the flight period for",sp_data_all$SPECIES[1],"at year", y,"weird predicted values")) | |
263 sp_data_all[, "FITTED"] <- NA | |
264 sp_data_all[, "COUNT_IMPUTED"] <- NA | |
265 sp_data_all[is.na(sp_data_all$COUNT), "COUNT_IMPUTED"] <- NA | |
266 sp_data_all[, "NM"] <- NA | |
267 } | |
268 else { | |
269 sp_data_all[, "COUNT_IMPUTED"] <- sp_data_all$COUNT | |
270 sp_data_all[is.na(sp_data_all$COUNT), "COUNT_IMPUTED"] <- sp_data_all$FITTED[is.na(sp_data_all$COUNT)] | |
271 # Define the flight curve from the fitted values and append them over | |
272 # years (this is one flight curve per year for all site) | |
273 site_sums <- aggregate(sp_data_all$FITTED, by = list(SITE = sp_data_all$SITE), | |
274 FUN = sum) | |
275 # Rename sum column | |
276 names(site_sums)[names(site_sums) == "x"] <- "SITE_YR_FSUM" | |
277 # Add data to sp_data data.frame (ensure merge does not sort the data!) | |
278 sp_data_all = merge(sp_data_all, site_sums, by <- c("SITE"), | |
279 all = TRUE, sort = FALSE) | |
280 # Calculate normalized values | |
281 sp_data_all[, "NM"] <- sp_data_all$FITTED/sp_data_all$SITE_YR_FSUM | |
282 } | |
283 } | |
284 } | |
285 else { | |
286 # Generate a list of values for all days from the additive model and use | |
287 # these value to fill the missing observations | |
288 sp_data_all[, "FITTED"] <- mgcv::predict.gam(gam_obj_site, newdata = sp_data_all[, | |
289 c("trimDAYNO", "SITE")], type = "response") | |
290 # force zeros at the beginning end end of the year | |
291 sp_data_all[sp_data_all$trimDAYNO < 60, "FITTED"] <- 0 | |
292 sp_data_all[sp_data_all$trimDAYNO > 305, "FITTED"] <- 0 | |
293 # if infinite number in predict replace with NA. | |
294 if(sum(is.infinite(sp_data_all[, "FITTED"]))>0){ | |
295 print(paste("Error in the flight period for",sp_data_all$SPECIES[1],"at year", y,"weird predicted values")) | |
296 sp_data_all[, "FITTED"] <- NA | |
297 sp_data_all[, "COUNT_IMPUTED"] <- NA | |
298 sp_data_all[is.na(sp_data_all$COUNT), "COUNT_IMPUTED"] <- NA | |
299 sp_data_all[, "NM"] <- NA | |
300 } | |
301 else { | |
302 sp_data_all[, "COUNT_IMPUTED"] <- sp_data_all$COUNT | |
303 sp_data_all[is.na(sp_data_all$COUNT), "COUNT_IMPUTED"] <- sp_data_all$FITTED[is.na(sp_data_all$COUNT)] | |
304 # Define the flight curve from the fitted values and append them over | |
305 # years (this is one flight curve per year for all site) | |
306 site_sums = aggregate(sp_data_all$FITTED, by = list(SITE = sp_data_all$SITE), | |
307 FUN = sum) | |
308 # Rename sum column | |
309 names(site_sums)[names(site_sums) == "x"] = "SITE_YR_FSUM" | |
310 # Add data to sp_data data.frame (ensure merge does not sort the data!) | |
311 sp_data_all = merge(sp_data_all, site_sums, by = c("SITE"), all = TRUE, | |
312 sort = FALSE) | |
313 # Calculate normalized values | |
314 sp_data_all[, "NM"] = sp_data_all$FITTED/sp_data_all$SITE_YR_FSUM | |
315 } | |
316 } | |
317 sp_data_filled <- sp_data_all | |
318 flight_curve <- data.frame(species = sp_data_filled$SPECIES, year = sp_data_filled$YEAR, | |
319 week = sp_data_filled$WEEK, DAYNO = sp_data_filled$DAYNO, DAYNO_adj = sp_data_filled$trimDAYNO, | |
320 nm = sp_data_filled$NM)[!duplicated(paste(sp_data_filled$YEAR, | |
321 sp_data_filled$DAYNO, sep = "_")), ] | |
322 flight_curve <- flight_curve[order(flight_curve$DAYNO), ] | |
323 # bind if exist else create | |
324 if (is.na(flight_curve$nm[1])) next() | |
325 | |
326 flight_pheno <- rbind(flight_pheno, flight_curve) | |
327 | |
328 } # end of year loop | |
329 } | |
330 else { | |
331 y <- unique(dataset$YEAR) | |
332 dataset_y <- dataset[dataset$YEAR == y, ] | |
333 # subset sites with enough visit and occurence | |
334 occ <- aggregate(dataset_y$COUNT, by = list(SITE = dataset_y$SITE), function(x) sum(x > 0)) | |
335 vis <- aggregate(dataset_y$COUNT, by = list(SITE = dataset_y$SITE), function(x) length(x)) | |
336 dataset_y <- dataset_y[dataset_y$SITE %in% occ$SITE[occ$x >= MinOccur], ] | |
337 dataset_y <- dataset_y[dataset_y$SITE %in% vis$SITE[vis$x >= MinVisit], ] | |
338 nsite <- length(unique(dataset_y$SITE)) | |
339 if (nsite < 1) { | |
340 stop(paste("No sites with sufficient visits and occurence, MinOccur:", MinOccur, " MinVisit: ", MinVisit, " for " ,dataset$SPECIES[1],"at year", y)) | |
341 } | |
342 # Determine missing days and add to dataset | |
343 sp_data_all <- year_day_func(dataset_y) | |
344 if (nsite > 200) { | |
345 sp_data_all <- sp_data_all[as.character(sp_data_all$SITE) %in% as.character(unique(dataset_y$SITE)[sample(1:nsite, | |
346 200, replace = F)]), ] | |
347 } | |
348 else { | |
349 sp_data_all <- sp_data_all | |
350 } | |
351 sp_data_all$trimDAYNO <- sp_data_all$DAYNO - min(sp_data_all$DAYNO) + 1 | |
352 print(paste("Fitting the GAM for",sp_data_all$SPECIES[1],"at year", y,":",Sys.time())) | |
353 if(length(unique(sp_data_all$SITE))>1){ | |
354 gam_obj_site <- try(mgcv::gam(COUNT ~ s(trimDAYNO, bs = "cr") + as.factor(SITE) -1, | |
355 data = sp_data_all, family = GamFamily), silent = TRUE) | |
356 } | |
357 else { | |
358 gam_obj_site <- try(mgcv::gam(COUNT ~ s(trimDAYNO, bs = "cr") -1, | |
359 data = sp_data_all, family = GamFamily), silent = TRUE) | |
360 } | |
361 # Give a second try if the GAM does not converge the first time | |
362 if (class(gam_obj_site)[1] == "try-error") { | |
363 # Determine missing days and add to dataset | |
364 sp_data_all <- year_day_func(dataset_y) | |
365 if (nsite > 200) { | |
366 sp_data_all <- sp_data_all[as.character(sp_data_all$SITE) %in% as.character(unique(dataset_y$SITE)[sample(1:nsite, | |
367 200, replace = F)]), ] | |
368 } | |
369 else { | |
370 sp_data_all <- sp_data_all | |
371 } | |
372 sp_data_all$trimDAYNO <- sp_data_all$DAYNO - min(sp_data_all$DAYNO) + 1 | |
373 print(paste("Fitting the GAM for",sp_data_all$SPECIES[1],"at year", y,"with",length(unique(sp_data_all$SITE)),"sites :",Sys.time(),"second try")) | |
374 if(length(unique(sp_data_all$SITE))>1){ | |
375 gam_obj_site <- try(mgcv::bam(COUNT ~ s(trimDAYNO, bs = "cr") + as.factor(SITE) - 1, | |
376 data = sp_data_all, family = GamFamily), silent = TRUE) | |
377 } | |
378 else { | |
379 gam_obj_site <- try(mgcv::gam(COUNT ~ s(trimDAYNO, bs = "cr") -1, | |
380 data = sp_data_all, family = GamFamily), silent = TRUE) | |
381 } | |
382 if (class(gam_obj_site)[1] == "try-error") { | |
383 print(paste("Error in fitting the flight period for",sp_data_all$SPECIES[1],"at year", y,"no convergence after two trial")) | |
384 sp_data_all[, "FITTED"] <- NA | |
385 sp_data_all[, "COUNT_IMPUTED"] <- NA | |
386 sp_data_all[is.na(sp_data_all$COUNT), "COUNT_IMPUTED"] <- NA | |
387 sp_data_all[, "NM"] <- NA | |
388 } | |
389 else { | |
390 # Generate a list of values for all days from the additive model and use | |
391 # these value to fill the missing observations | |
392 sp_data_all[, "FITTED"] <- mgcv::predict.gam(gam_obj_site, newdata = sp_data_all[, | |
393 c("trimDAYNO", "SITE")], type = "response") | |
394 # force zeros at the beginning end end of the year | |
395 sp_data_all[sp_data_all$trimDAYNO < 60, "FITTED"] <- 0 | |
396 sp_data_all[sp_data_all$trimDAYNO > 305, "FITTED"] <- 0 | |
397 # if infinite number in predict replace with NA. | |
398 if(sum(is.infinite(sp_data_all[, "FITTED"]))>0){ | |
399 print(paste("Error in the flight period for",sp_data_all$SPECIES[1],"at year", y,"weird predicted values")) | |
400 sp_data_all[, "FITTED"] <- NA | |
401 sp_data_all[, "COUNT_IMPUTED"] <- NA | |
402 sp_data_all[is.na(sp_data_all$COUNT), "COUNT_IMPUTED"] <- NA | |
403 sp_data_all[, "NM"] <- NA | |
404 } | |
405 else { | |
406 sp_data_all[, "COUNT_IMPUTED"] <- sp_data_all$COUNT | |
407 sp_data_all[is.na(sp_data_all$COUNT), "COUNT_IMPUTED"] <- sp_data_all$FITTED[is.na(sp_data_all$COUNT)] | |
408 # Define the flight curve from the fitted values and append them over | |
409 # years (this is one flight curve per year for all site) | |
410 site_sums <- aggregate(sp_data_all$FITTED, by = list(SITE = sp_data_all$SITE), | |
411 FUN = sum) | |
412 # Rename sum column | |
413 names(site_sums)[names(site_sums) == "x"] <- "SITE_YR_FSUM" | |
414 # Add data to sp_data data.frame (ensure merge does not sort the data!) | |
415 sp_data_all = merge(sp_data_all, site_sums, by <- c("SITE"), | |
416 all = TRUE, sort = FALSE) | |
417 # Calculate normalized values | |
418 sp_data_all[, "NM"] <- sp_data_all$FITTED/sp_data_all$SITE_YR_FSUM | |
419 } | |
420 } | |
421 } | |
422 else { | |
423 # Generate a list of values for all days from the additive model and use | |
424 # these value to fill the missing observations | |
425 sp_data_all[, "FITTED"] <- mgcv::predict.gam(gam_obj_site, newdata = sp_data_all[, | |
426 c("trimDAYNO", "SITE")], type = "response") | |
427 # force zeros at the beginning end end of the year | |
428 sp_data_all[sp_data_all$trimDAYNO < 60, "FITTED"] <- 0 | |
429 sp_data_all[sp_data_all$trimDAYNO > 305, "FITTED"] <- 0 | |
430 # if infinite number in predict replace with NA. | |
431 if(sum(is.infinite(sp_data_all[, "FITTED"]))>0){ | |
432 print(paste("Error in the flight period for",sp_data_all$SPECIES[1],"at year", y,"weird predicted values")) | |
433 sp_data_all[, "FITTED"] <- NA | |
434 sp_data_all[, "COUNT_IMPUTED"] <- NA | |
435 sp_data_all[is.na(sp_data_all$COUNT), "COUNT_IMPUTED"] <- NA | |
436 sp_data_all[, "NM"] <- NA | |
437 } | |
438 else { | |
439 sp_data_all[, "COUNT_IMPUTED"] <- sp_data_all$COUNT | |
440 sp_data_all[is.na(sp_data_all$COUNT), "COUNT_IMPUTED"] <- sp_data_all$FITTED[is.na(sp_data_all$COUNT)] | |
441 # Define the flight curve from the fitted values and append them over | |
442 # years (this is one flight curve per year for all site) | |
443 site_sums = aggregate(sp_data_all$FITTED, by = list(SITE = sp_data_all$SITE), | |
444 FUN = sum) | |
445 # Rename sum column | |
446 names(site_sums)[names(site_sums) == "x"] = "SITE_YR_FSUM" | |
447 # Add data to sp_data data.frame (ensure merge does not sort the data!) | |
448 sp_data_all = merge(sp_data_all, site_sums, by = c("SITE"), all = TRUE, | |
449 sort = FALSE) | |
450 # Calculate normalized values | |
451 sp_data_all[, "NM"] = sp_data_all$FITTED/sp_data_all$SITE_YR_FSUM | |
452 } | |
453 } | |
454 sp_data_filled <- sp_data_all | |
455 flight_curve <- data.frame(species = sp_data_filled$SPECIES, year = sp_data_filled$YEAR, | |
456 week = sp_data_filled$WEEK, DAYNO = sp_data_filled$DAYNO, DAYNO_adj = sp_data_filled$trimDAYNO, | |
457 nm = sp_data_filled$NM)[!duplicated(paste(sp_data_filled$YEAR, | |
458 sp_data_filled$DAYNO, sep = "_")), ] | |
459 flight_curve <- flight_curve[order(flight_curve$DAYNO), ] | |
460 | |
461 flight_pheno <- rbind(flight_pheno, flight_curve) | |
462 | |
463 } | |
464 return(flight_pheno) | |
465 } | |
466 | |
467 | |
468 #' abundance_index Function | |
469 #' | |
470 #' This function compute the Abundance Index across sites and years from your dataset and the regional flight curve | |
471 #' @param your_dataset A data.frame containing original species count data. The data format is a csv (comma "," separated) with 6 columns with headers, namely "species","transect_id","visit_year","visit_month","visit_day","sp_count" | |
472 #' @param flight_pheno A data.frame for the regional flight curve computed with the function flight_curve | |
473 #' @keywords standardize flight curve | |
474 #' @export | |
475 #' @examples | |
476 #' abundance_index() | |
477 | |
478 abundance_index <- function(your_dataset,flight_pheno) { | |
479 | |
480 your_dataset$DAYNO <- strptime(paste(your_dataset$DAY, your_dataset$MONTH, | |
481 your_dataset$YEAR, sep = "/"), "%d/%m/%Y")$yday + 1 | |
482 | |
483 dataset <- your_dataset[, c("SPECIES", "SITE", "YEAR", "MONTH", | |
484 "DAY", "DAYNO", "COUNT")] | |
485 | |
486 sample_year <- unique(dataset$YEAR) | |
487 sample_year <- sample_year[order(sample_year)] | |
488 | |
489 | |
490 if (length(sample_year)>1){ | |
491 | |
492 for (y in sample_year) { | |
493 | |
494 year_pheno <- flight_pheno[flight_pheno$year == y, ] | |
495 | |
496 dataset_y <- dataset[dataset$YEAR == y, ] | |
497 | |
498 sp_data_site <- year_day_func(dataset_y) | |
499 sp_data_site$trimDAYNO <- sp_data_site$DAYNO - min(sp_data_site$DAYNO) + 1 | |
500 | |
501 sp_data_site <- merge(sp_data_site, year_pheno[, c("DAYNO", "nm")], | |
502 by = c("DAYNO"), all.x = TRUE, sort = FALSE) | |
503 | |
504 # compute proportion of the flight curve sampled due to missing visits | |
505 pro_missing_count <- data.frame(SITE = sp_data_site$SITE, WEEK = sp_data_site$WEEK, | |
506 NM = sp_data_site$nm, COUNT = sp_data_site$COUNT, ANCHOR = sp_data_site$ANCHOR) | |
507 pro_missing_count$site_week <- paste(pro_missing_count$SITE, pro_missing_count$WEEK, | |
508 sep = "_") | |
509 siteweeknocount <- aggregate(pro_missing_count$COUNT, by = list(pro_missing_count$site_week), | |
510 function(x) sum(!is.na(x)) == 0) | |
511 pro_missing_count <- pro_missing_count[pro_missing_count$site_week %in% | |
512 siteweeknocount$Group.1[siteweeknocount$x == TRUE], ] | |
513 pro_count_agg <- aggregate(pro_missing_count$NM, by = list(pro_missing_count$SITE), | |
514 function(x) 1 - sum(x, na.rm = T)) | |
515 names(pro_count_agg) <- c("SITE", "PROP_PHENO_SAMPLED") | |
516 | |
517 # remove samples outside the monitoring window | |
518 sp_data_site$COUNT[sp_data_site$nm==0] <- NA | |
519 | |
520 # Compute the regional GAM index | |
521 | |
522 if(length(unique(sp_data_site$SITE))>1){ | |
523 glm_obj_site <- glm(COUNT ~ factor(SITE) + offset(log(nm)) - 1, data = sp_data_site, | |
524 family = quasipoisson(link = "log"), control = list(maxit = 100)) | |
525 } else { | |
526 glm_obj_site <- glm(COUNT ~ offset(log(nm)) - 1, data = sp_data_site, | |
527 family = quasipoisson(link = "log"), control = list(maxit = 100)) | |
528 } | |
529 | |
530 sp_data_site[, "FITTED"] <- predict.glm(glm_obj_site, newdata = sp_data_site, | |
531 type = "response") | |
532 sp_data_site[, "COUNT_IMPUTED"] <- sp_data_site$COUNT | |
533 sp_data_site[is.na(sp_data_site$COUNT), "COUNT_IMPUTED"] <- sp_data_site$FITTED[is.na(sp_data_site$COUNT)] | |
534 | |
535 ## add fitted value for missing mid-week data | |
536 sp_data_site <- sp_data_site[!paste(sp_data_site$DAY_WEEK, sp_data_site$COUNT) %in% | |
537 c("1 NA", "2 NA", "3 NA", "5 NA", "6 NA", "7 NA"), ] | |
538 | |
539 ## remove all added mid-week values for weeks with real count | |
540 ## (observation) | |
541 sp_data_site$site_week <- paste(sp_data_site$SITE, sp_data_site$WEEK, | |
542 sep = "_") | |
543 siteweekcount <- aggregate(sp_data_site$COUNT, by = list(sp_data_site$site_week), | |
544 function(x) sum(!is.na(x)) > 0) | |
545 sp_data_site <- sp_data_site[!(is.na(sp_data_site$COUNT) & (sp_data_site$site_week %in% | |
546 siteweekcount$Group.1[siteweekcount$x == TRUE])), names(sp_data_site) != | |
547 "site_week"] | |
548 | |
549 ## Compute the regional GAM index | |
550 print(paste("Compute index for",sp_data_site$SPECIES[1],"at year", y,"for",length(unique(sp_data_site$SITE)),"sites:",Sys.time())) | |
551 regional_gam_index <- trap_index(sp_data_site, data_col = "COUNT_IMPUTED", | |
552 time_col = "DAYNO", by_col = c("SPECIES", "SITE", "YEAR")) | |
553 | |
554 cumu_index <- merge(regional_gam_index, pro_count_agg, by = c("SITE"), | |
555 all.x = TRUE, sort = FALSE) | |
556 names(cumu_index) <- c("SITE", "SPECIES", "YEAR", "regional_gam", "prop_pheno_sampled") | |
557 | |
558 cumu_index <- cumu_index[order(cumu_index$SITE), ] | |
559 | |
560 # bind if exist else create | |
561 if ("cumullated_indices" %in% ls()) { | |
562 cumullated_indices <- rbind(cumullated_indices, cumu_index) | |
563 } else { | |
564 cumullated_indices <- cumu_index | |
565 } | |
566 | |
567 } # end of year loop | |
568 | |
569 } else { | |
570 | |
571 y <- unique(dataset$YEAR) | |
572 year_pheno <- flight_pheno[flight_pheno$year == y, ] | |
573 | |
574 dataset_y <- dataset[dataset$YEAR == y, ] | |
575 | |
576 sp_data_site <- year_day_func(dataset_y) | |
577 sp_data_site$trimDAYNO <- sp_data_site$DAYNO - min(sp_data_site$DAYNO) + 1 | |
578 | |
579 sp_data_site <- merge(sp_data_site, year_pheno[, c("DAYNO", "nm")], | |
580 by = c("DAYNO"), all.x = TRUE, sort = FALSE) | |
581 | |
582 # compute proportion of the flight curve sampled due to missing visits | |
583 pro_missing_count <- data.frame(SITE = sp_data_site$SITE, WEEK = sp_data_site$WEEK, | |
584 NM = sp_data_site$nm, COUNT = sp_data_site$COUNT, ANCHOR = sp_data_site$ANCHOR) | |
585 pro_missing_count$site_week <- paste(pro_missing_count$SITE, pro_missing_count$WEEK, | |
586 sep = "_") | |
587 siteweeknocount <- aggregate(pro_missing_count$COUNT, by = list(pro_missing_count$site_week), | |
588 function(x) sum(!is.na(x)) == 0) | |
589 pro_missing_count <- pro_missing_count[pro_missing_count$site_week %in% | |
590 siteweeknocount$Group.1[siteweeknocount$x == TRUE], ] | |
591 pro_count_agg <- aggregate(pro_missing_count$NM, by = list(pro_missing_count$SITE), | |
592 function(x) 1 - sum(x, na.rm = T)) | |
593 names(pro_count_agg) <- c("SITE", "PROP_PHENO_SAMPLED") | |
594 | |
595 # remove samples outside the monitoring window | |
596 sp_data_site$COUNT[sp_data_site$nm==0] <- NA | |
597 | |
598 # Compute the regional GAM index | |
599 if(length(unique(sp_data_site$SITE))>1){ | |
600 glm_obj_site <- glm(COUNT ~ factor(SITE) + offset(log(nm)) - 1, data = sp_data_site, | |
601 family = quasipoisson(link = "log"), control = list(maxit = 100)) | |
602 } else { | |
603 glm_obj_site <- glm(COUNT ~ offset(log(nm)) - 1, data = sp_data_site, | |
604 family = quasipoisson(link = "log"), control = list(maxit = 100)) | |
605 } | |
606 | |
607 sp_data_site[, "FITTED"] <- predict.glm(glm_obj_site, newdata = sp_data_site, | |
608 type = "response") | |
609 sp_data_site[, "COUNT_IMPUTED"] <- sp_data_site$COUNT | |
610 sp_data_site[is.na(sp_data_site$COUNT), "COUNT_IMPUTED"] <- sp_data_site$FITTED[is.na(sp_data_site$COUNT)] | |
611 | |
612 # add fitted value for missing mid-week data | |
613 sp_data_site <- sp_data_site[!paste(sp_data_site$DAY_WEEK, sp_data_site$COUNT) %in% | |
614 c("1 NA", "2 NA", "3 NA", "5 NA", "6 NA", "7 NA"), ] | |
615 | |
616 # remove all added mid-week values for weeks with real count | |
617 # (observation) | |
618 sp_data_site$site_week <- paste(sp_data_site$SITE, sp_data_site$WEEK, | |
619 sep = "_") | |
620 siteweekcount <- aggregate(sp_data_site$COUNT, by = list(sp_data_site$site_week), | |
621 function(x) sum(!is.na(x)) > 0) | |
622 sp_data_site <- sp_data_site[!(is.na(sp_data_site$COUNT) & (sp_data_site$site_week %in% | |
623 siteweekcount$Group.1[siteweekcount$x == TRUE])), names(sp_data_site) != | |
624 "site_week"] | |
625 | |
626 # Compute the regional gam index | |
627 print(paste("Compute index for",sp_data_site$SPECIES[1],"at year", y,"for",length(unique(sp_data_site$SITE)),"sites:",Sys.time())) | |
628 regional_gam_index <- trap_index(sp_data_site, data_col = "COUNT_IMPUTED", | |
629 time_col = "DAYNO", by_col = c("SPECIES", "SITE", "YEAR")) | |
630 | |
631 cumu_index <- merge(regional_gam_index, pro_count_agg, by = c("SITE"), | |
632 all.x = TRUE, sort = FALSE) | |
633 names(cumu_index) <- c("SITE", "SPECIES", "YEAR", "regional_gam", "prop_pheno_sampled") | |
634 | |
635 cumu_index <- cumu_index[order(cumu_index$SITE), ] | |
636 | |
637 # bind if exist else create | |
638 if ("cumullated_indices" %in% ls()) { | |
639 cumullated_indices <- rbind(cumullated_indices, cumu_index) | |
640 } else { | |
641 cumullated_indices <- cumu_index | |
642 } | |
643 | |
644 } | |
645 | |
646 return(cumullated_indices) | |
647 | |
648 } |