-
Notifications
You must be signed in to change notification settings - Fork 0
/
MacArthur_Paper_Analysis_Tanzania_Thresh10_Soc_RECREATE.R
623 lines (520 loc) · 22.7 KB
/
MacArthur_Paper_Analysis_Tanzania_Thresh10_Soc_RECREATE.R
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
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
library(sp)
library(jsonlite)
library(xtable)
library(stargazer)
library(RCurl)
library(maptools)
library(geosphere)
library(ncf)
library(gridExtra)
require(grid)
library(multiwayvcov)
library(lmtest)
library(reshape2)
#---------------------------------------------------#
#Settings
#---------------------------------------------------#
forest_thresh = 10
restrict_analysis = FALSE
#---------------------------------------------------#
#Download Data for Analysis - This step may take a while (up to hours)
#---------------------------------------------------#
#source("RDownload.R")
#mDir = getwd()
active_dir_path <- "/Users/rbtrichler/Box Sync/MacArthur"
#---------------------------------------------------#
#Load the dataframes in for analysis after download.
#---------------------------------------------------#
csv <- paste(active_dir_path, "/extracts/africa.csv", sep="")
json <- paste(active_dir_path, "/extracts/africa.json", sep="")
dta <- read.csv(csv)
vars <- fromJSON(txt=json)
#--------------------------------------------------#
#Subset the Cell Dataframe
#--------------------------------------------------#
dta2 <- dta[dta$NAME_0 == "Tanzania",]
#--------------------------------------------------#
#Load and Subset the Cell Spatial Datframe
#--------------------------------------------------#
spdf_cells <- paste(active_dir_path, "/grids/africa_grid.shp", sep="")
cells <- readShapePoly(spdf_cells)
#Keep only relevant cells
AOI_cells <- sp::merge(cells, dta2, by="ID", all.x=FALSE)
if(restrict_analysis != FALSE)
{
AOI_cells<- AOI_cells[1:restrict_analysis,]
}
proj4string(AOI_cells) <- CRS("+proj=longlat +ellps=WGS84 +datum=WGS84")
#--------------------------------------------------#
#Load and Subset the Spatial ADM zone data
#--------------------------------------------------#
spdf_adm_path <- paste(active_dir_path, "/ADM2/GADM_MacEcohotspotSubset_ADM2.shp", sep="")
spdf_adm <- readShapePoly(spdf_adm_path)
#Keep only relevant ADM data
spdf_adm <- spdf_adm[spdf_adm@data$NAME_0 == "Tanzania",]
#--------------------------------------------------#
#Load and Subset the MacArthur Aid Data
#--------------------------------------------------#
location_csv <- paste(active_dir_path, "/MacArthur_Geocoded_data/locations.csv", sep="")
locations <- read.csv(location_csv)
locations2 <- locations[grep("Tanzania", locations$gazetteer_adm_name),]
coords = cbind(locations2$longitude, locations2$latitude)
#--------------------------------------------------#
#Create a Spatial Dataframe of the MacArthur Aid Data
#--------------------------------------------------#
Mac_spdf <- SpatialPointsDataFrame(coords, locations2)
proj4string(Mac_spdf) <- CRS("+proj=longlat +ellps=WGS84 +datum=WGS84")
#Join in the year of aid allocation from projects_ancillary
Mac_years <- paste(active_dir_path, "/MacArthur_Geocoded_data/projects_ancillary.csv", sep="")
Mac_yr_dta <- read.csv(Mac_years)
Mac_spdf <- merge(Mac_spdf, Mac_yr_dta, by="project_id")
#Subset by precision code = 1 or 2
Mac_prec <- Mac_spdf[Mac_spdf@data$precision_code<=2,]
Mac_spdf <- Mac_prec
#Subset by sector code (combined social projects)
Mac_sector <- Mac_spdf[Mac_spdf@data$crs_sector_code%in%c("110","120","130"),]
Mac_spdf <- Mac_sector
#Subset by status = implementation or completion (not pipeline)
Mac_status <- Mac_spdf[Mac_spdf@data$status_code%in%c("2","3","11"),]
Mac_spdf <- Mac_status
write.csv(Mac_spdf@data,"/Users/rbtrichler/Box Sync/MacArthur/Mac_spdf_TanzSoc.csv")
#write.csv(Mac_spdf@data,"/home/aiddata/Desktop/Github/MacArthur/modelData/Mac_spdf_TanzaniaInfra.csv")
#writePointsShape(Mac_spdf, "/home/aiddata/Desktop/Github/MacArthur/modelData/Mac_spdf_Tanzania.shp")
#--------------------------------------------------#
#Create threshdolded forest datasets
#--------------------------------------------------#
ndviDTA <- AOI_cells#dta2[c("tc00_e", "lnyx_2000e")]
ndviDTA_for <- ndviDTA[ndviDTA$tc00_e >= forest_thresh,]
ndviDTA_notfor <- ndviDTA[ndviDTA$tc00_e <= forest_thresh,]
AOI_cells = ndviDTA_for
#--------------------------------------------------#
#Calculate the distances between cells and MacArthur points
#--------------------------------------------------#
source("RDist.R")
dMatrix <- RDist(AOI_cells, Mac_spdf)
#In dMatrix, every column is a cell (referenced in order to AOI_cells)
#Every row is a MacArthur project (referenced in order to Mac_spdf)
#Average distance in KM:
avgDistKm <- mean(dMatrix) / 1000
#Minimum
col_mins <- do.call(pmin, lapply(1:nrow(dMatrix), function(i)dMatrix[i,]))
minDistKm <- mean(col_mins) / 1000
#--------------------------------------------------#
#Calculate the correlogram
#--------------------------------------------------#
#correlogram_data <- correlog(x = coordinates(AOI_cells)[,1], y = coordinates(AOI_cells)[,2], z=AOI_cells$lnyx_1999e, increment=5, latlon=TRUE, na.rm=TRUE, resamp=5)
#save (correlogram_data, file="/home/aiddata/Desktop/Github/MacArthur/modelData/tanzania_correl.RData")
load("modelData/tanzania_correl.RData")
#save data into a function to calculate the distance-decay penalty later.
#Chinese projects are "weighted" according to their distance.
#The absolulute correlation for a given distance is used as a weight.
#Projects at distances with a higher positive or inverse correlation are given the highest
#weights.
dVals <- abs(correlogram_data$mean.of.class)
cVals <- abs(correlogram_data$correlation)
#--------------------------------------------------#
#Calculate the average spatial decays for visualization
#--------------------------------------------------#
#no weights, all before the first time X = 0 are counted.
thresh_dMatrix <- dMatrix
#thresh_dMatrix[thresh_dMatrix > (as.numeric(correlogram_data$x.intercept)*1000)] <- NA
thresh_dMatrix[thresh_dMatrix >= 100000] <- NA
total_distance_km <- colSums(thresh_dMatrix, na.rm=TRUE) / 1000
AOI_cells$thresh_tot_proj <- apply(thresh_dMatrix, 2, function(x) length(which(!is.na(x))))
AOI_cells$thresh_totDist <- total_distance_km
AOI_cells$thresh_avgDist <- AOI_cells$thresh_totDist / AOI_cells$thresh_tot_proj
#distance decay
decay_dMatrix <- dMatrix
decay_dMatrix_adj <- apply(decay_dMatrix, 1:2, function(x){(cVals[which.min(abs(dVals - x))])[[1]]})
AOI_cells$thresh_weightedDist <- colSums(decay_dMatrix_adj) / 1000
# #Drop cells which have no projects within the threshold
# AOI_cells2 <- AOI_cells[!(is.na(AOI_cells@data$thresh_avgDist)),]
# AOI_cellsBack <- AOI_cells
# AOI_cells <- AOI_cells2
#writePolyShape(AOI_cells, "/home/aiddata/Desktop/Github/MacArthur/modelData/AOI_cells_Tanzania.shp")
#--------------------------------------------------#
#Calculate the over-time treatment effects
#--------------------------------------------------#
all_years <- unique(Mac_spdf$year)
all_years <- all_years[!is.na(all_years)]
record_length <- c(2001:2014)
dYears <- list()
#Drop all MacArthur projects that have no start or end date.
Mac_spdf <- Mac_spdf[!is.na(Mac_spdf$year),]
for(years in 1:length(record_length))
{
if(record_length[[years]] %in% all_years)
{
year <- record_length[[years]]
ThisYearMac <- Mac_spdf[Mac_spdf@data$year == year,]
dYears[[years]] <- RDist(AOI_cells, ThisYearMac)
}
else
{
dYears[[years]] <- 0
}
}
#In dMatrix, every column is a cell (referenced in order to AOI_cells)
#Every row is a MacArthur project for that year (referenced in order to Mac_spdf)
AvgYears <- vector()
for(years in 1:length(record_length))
{
if(record_length[[years]] %in% all_years)
{
AvgYears[[years]] <- mean(dYears[[years]]) / 1000
}
else
{
AvgYears[[years]] <- 0
}
}
Avg_MinYears <- vector()
for(years in 1:length(record_length))
{
if(record_length[[years]] %in% all_years)
{
col_mins_year <- do.call(pmin, lapply(1:nrow(dYears[[years]]), function(i)dYears[[years]][i,]))
nameRef <- paste("MinYr_",record_length[[years]], sep="")
AOI_cells@data[nameRef] <- col_mins_year / 1000
Avg_MinYears[[years]] <- mean(col_mins_year) / 1000
}
else
{
Avg_MinYears[[years]] <- 0
nameRef <- paste("MinYr_",record_length[[years]], sep="")
AOI_cells@data[nameRef] <- 0
}
}
Dist_Decay_Yrs <- vector()
dvz <- cVals
dvz[dVals > correlogram_data$x.intercept] <- 0
for(years in 1:length(record_length))
{
if(record_length[[years]] %in% all_years)
{
t_dyears <- dYears[[years]] / 1000
decay_dMatrix_adj <- apply(t_dyears, 1:2, function(x){(dvz[which.min(abs(dVals - x))])[[1]]})
nameRef <- paste("DecayYr_",record_length[[years]], sep="")
AOI_cells@data[nameRef] <- colSums(decay_dMatrix_adj)
Dist_Decay_Yrs[[years]] <- sum(colSums(decay_dMatrix_adj))
}
else
{
Dist_Decay_Yrs[[years]] <- 0
nameRef <- paste("DecayYr_",record_length[[years]], sep="")
AOI_cells@data[nameRef] <- 0
}
}
#---------------------------
#PROJECT YEARS START COUNT (count of all projects within 100km)
#---------------------------
Proj_Thresh_Count_Yrs <- vector()
#in KM
thresh <- 100
cthreshVals <- cVals
cthreshVals[dVals <= thresh] <- 1
cthreshVals[dVals > thresh] <- 0
for(years in 1:length(record_length))
{
if(record_length[[years]] %in% all_years)
{
t_dyears <- dYears[[years]] / 1000
decay_dMatrix_adj <- apply(t_dyears, 1:2, function(x){(cthreshVals[which.min(abs(dVals - x))])[[1]]})
nameRef <- paste("Proj_Thresh_Count_Yrs_",record_length[[years]], sep="")
AOI_cells@data[nameRef] <- colSums(decay_dMatrix_adj)
Proj_Thresh_Count_Yrs[[years]] <- sum(colSums(decay_dMatrix_adj))
}
else
{
Proj_Thresh_Count_Yrs[[years]] <- 0
nameRef <- paste("Proj_Thresh_Count_Yrs_",record_length[[years]], sep="")
AOI_cells@data[nameRef] <- 0
}
}
#---------------------------
#Limit distance decay threshold to 100km (rather than the x-intercept)
#---------------------------
DistDecay100 <- vector()
#in KM
thresh <- 100
cthreshVals_decay100 <- cVals
cthreshVals_decay100[dVals > thresh] <- 0
for(years in 1:length(record_length))
{
if(record_length[[years]] %in% all_years)
{
t_dyears <- dYears[[years]] / 1000
decay_dMatrix_adj <- apply(t_dyears, 1:2, function(x){(cthreshVals_decay100[which.min(abs(dVals - x))])[[1]]})
nameRef <- paste("DistDecay100_",record_length[[years]], sep="")
AOI_cells@data[nameRef] <- colSums(decay_dMatrix_adj)
DistDecay100[[years]] <- sum(colSums(decay_dMatrix_adj))
}
else
{
DistDecay100[[years]] <- 0
nameRef <- paste("DistDecay100_",record_length[[years]], sep="")
AOI_cells@data[nameRef] <- 0
}
}
# CountProj_Years <- vector()
# for(years in 1:length(record_length))
# {
# if(record_length[[years]] %in% all_years)
# {
# CountProj_Years[[years]] <- nrow(dYears[[years]])
# }
# else
# {
# CountProj_Years[[years]] <- 0
# }
# }
#
# #Build a quick temporal dataframe for plotting and ordering
# TempDF <- cbind.data.frame(record_length, AvgYears, Avg_MinYears, CountProj_Years, Dist_Decay_Yrs, Proj_Thresh_Count_Yrs)
# TempDF <- TempDF[with(TempDF, order(TempDF[,1])),]
#--------------------------------------------------#
#Pre-processing for analysis
#--------------------------------------------------#
DFa <- AOI_cells@data
#Drop irrelevant variables:
dropvars <- c("XMIN","XMAX","YMIN","YMAX","OBJECTID","ID_0","ISO","NAME_0","HASC_2","ID_1","NAME_1","NAME_2",
"CCN_2","CCA_2","TYPE_2","ENGTYPE_2","NL_NAME_2","VARNAME_2","Shape_Leng","Shape_Area",
"thresh_tot_proj","thresh_totDist","thresh_avgDist","thresh_weightedDist")
DFa <- DFa[,!(names(DFa) %in% dropvars)]
DFa_hist <- DFa
DFa <- DFa[, -grep("(19)", names(DFa))]
#Prep for wide to long translation
substrRight <- function(x, n){
substr(x, nchar(x)-n+1, nchar(x))
}
rename_header <- function(x,sub)
{
t <- paste(substr(x, 1, 0), sub, substr(x, 1, nchar(x)), sep = "")
substr(t, 1, nchar(t)-1)
}
#names(DFa)[25:143]<- sapply(names(DFa)[25:143], function(x) {rename_header(x, substrRight(x,1))})
#names(DFa)[6:19] <- sapply(names(DFa)[6:19], function(x){substr(x, 5, nchar(x))})
#Reorder variables
for (i in 2:length(DFa)) {
if (substr(colnames(DFa)[i], 1, 4) == "at41"){
name = substr(colnames(DFa)[i],1, 4)
year = substr(colnames(DFa)[i], 6, 9)
letter = substr(colnames(DFa)[i], 10,10)
dt = paste(letter,name,"_",year,sep="")
colnames(DFa)[i] <- dt
}
}
for (i in 2:length(DFa)) {
if (substr(colnames(DFa)[i], 1, 4) == "pc41"){
name = substr(colnames(DFa)[i],1, 4)
year = substr(colnames(DFa)[i], 6, 9)
letter = substr(colnames(DFa)[i], 10,10)
dt = paste(letter,name,"_",year,sep="")
colnames(DFa)[i] <- dt
}
}
for (i in 2:length(DFa)) {
if (substr(colnames(DFa)[i], 1, 4) == "ncc4"){
name = substr(colnames(DFa)[i],1, 4)
year = substr(colnames(DFa)[i], 6, 9)
letter = substr(colnames(DFa)[i], 10,10)
dt = paste(letter,name,"_",year,sep="")
colnames(DFa)[i] <- dt
}
}
for (i in 2:length(DFa)) {
if (substr(colnames(DFa)[i], 1, 4) == "lnyx"){
name = substr(colnames(DFa)[i],1, 4)
year = substr(colnames(DFa)[i], 6, 9)
letter = substr(colnames(DFa)[i], 10,10)
dt = paste(letter,name,"_",year,sep="")
colnames(DFa)[i] <- dt
}
}
for (i in 2:length(DFa))
{
colnames(DFa)[i] <- sub("per_loss_","loss_",colnames(DFa)[i])
}
for (i in 2:length(DFa))
{
colnames(DFa)[i] <- sub("Proj_Thresh_Count_Yrs","ProjCnt100",colnames(DFa)[i])
}
#drop data for the year 2000, 2013-2015 (NTL ends in 2012)
DFa2 <- DFa[, -grep("(2000)", names(DFa))]
# DFa3 <- DFa2[, -grep("(2013)", names(DFa2))]
#DFa <- DFa[, -grep("(2014)", names(DFa3))]
#DFa <- DFa4
DFa3 <- DFa2[, -grep("^encc4", names(DFa2))]
PCloss <- grep("^loss", names(DFa3))
mean_ln <- grep("^elnyx", names(DFa3))
NTL <- grep("^encc4", names(DFa3))
minairTemp <- grep("^mat41", names(DFa3))
maxairTemp <- grep("^xat41", names(DFa3))
meanairTemp <- grep("^eat41", names(DFa3))
minPre <- grep("^mpc41", names(DFa3))
maxPre <- grep("^xpc41", names(DFa3))
meanPre <- grep("^epc41", names(DFa3))
MinDist <- grep("^MinYr", names(DFa3))
DecayDist <- grep("^DecayYr", names(DFa3))
ProjCount <- grep("^ProjCnt100", names(DFa3))
DecayDist100<- grep("^DistDecay100", names(DFa3))
#--------------------------------------------------#
#Selection of temporally-varying variables and shift from wide- to long-form
#--------------------------------------------------#
all_reshape <- c(PCloss, mean_ln, minairTemp, maxairTemp, meanairTemp, minPre, maxPre, meanPre, MinDist, DecayDist, ProjCount, DecayDist100)
DFa4 <- reshape(DFa3, varying=all_reshape,direction="long", idvar="ID", sep="_", timevar="Year")
DFa <- DFa4
#Rename names to something interpretable...
names(DFa)[names(DFa) == "ID_2"] = "District"
names(DFa)[names(DFa) == "loss"] = "Forest_Loss"
names(DFa)[names(DFa) == "encc4"] = "NighttimeLights"
names(DFa)[names(DFa) == "mat41"] = "MinTemp"
names(DFa)[names(DFa) == "xat41"] = "MaxTemp"
names(DFa)[names(DFa) == "eat41"] = "MeanTemp"
names(DFa)[names(DFa) == "mpc41"] = "MinPrecip"
names(DFa)[names(DFa) == "xpc41"] = "MaxPrecip"
names(DFa)[names(DFa) == "epc41"] = "MeanPrecip"
names(DFa)[names(DFa) == "selv_e"] = "Elevation"
names(DFa)[names(DFa) == "sslp_e"] = "Slope"
names(DFa)[names(DFa) == "dari_e"] = "RivDist"
names(DFa)[names(DFa) == "droa_e"] = "RoadDist"
names(DFa)[names(DFa) == "am50_e"] = "UrbTravTime"
names(DFa)[names(DFa) == "DistDecay100"] = "DecayYr100"
#--------------------------------------------------#
#Additive Year-on-Year
#--------------------------------------------------#
Panel_Data <- DFa
#Panel_Data$MinYr_additive <- NA
Panel_Data$DecayYr_additive <- NA
Panel_Data$Forest_Loss_additive <- NA
Panel_Data$DecayYr100_additive <- NA
Panel_Data$ProjCnt100_additive <- NA
Panel_Data <- Panel_Data[with(Panel_Data, order(Panel_Data["ID"], Panel_Data["Year"])),]
calc_add <- function(Fdta, year, ID, var)
{
a.dta <- Fdta[Fdta$ID == ID,]
b.dta <- a.dta[a.dta$Year <= year,]
exec_st <- paste("sum(b.dta$",var,")",sep="")
return(eval(parse(text=exec_st)))
}
for(i in 1:length(Panel_Data[[1]]))
{
Panel_Data["DecayYr_additive"][i,] <- calc_add(Panel_Data, Panel_Data[i,]["Year"][[1]], Panel_Data[i,]["ID"][[1]], "DecayYr")
Panel_Data["DecayYr100_additive"][i,] <- calc_add(Panel_Data, Panel_Data[i,]["Year"][[1]], Panel_Data[i,]["ID"][[1]], "DecayYr100")
#Panel_Data["MinYr_additive"][i,] <- calc_add(Panel_Data, Panel_Data[i,]["Year"][[1]], Panel_Data[i,]["ID"][[1]], "MinYr")
Panel_Data["Forest_Loss_additive"][i,] <- calc_add(Panel_Data, Panel_Data[i,]["Year"][[1]], Panel_Data[i,]["ID"][[1]], "Forest_Loss")
Panel_Data["ProjCnt100_additive"][i,] <- calc_add(Panel_Data, Panel_Data[i,]["Year"][[1]], Panel_Data[i,]["ID"][[1]], "ProjCnt100")
}
#Control Variables
val_lookup <- function(dta, var, id)
{
ret_find_exec <- paste("dta$",var)
}
pre_trend_func <- function(dta, id)
{
#build a dataframe
NDVI_reshape <- c("lnyx_1990e", "lnyx_1991e", "lnyx_1992e", "lnyx_1993e", "lnyx_1994e", "lnyx_1995e", "lnyx_1996e", "lnyx_1997e", "lnyx_1998e", "lnyx_1999e")
DFa_mdl <- DFa_hist[DFa_hist$ID == id,]
mdl_dta <- reshape(DFa_mdl, varying=NDVI_reshape,direction="long", idvar="ID", sep="_", timevar="Year")
return(lm(lnyx ~ Year, data=mdl_dta)$coefficients["Year"][[1]])
}
Panel_Data["DecayAddControl"] <- NA
Panel_Data["DecayAddControl100"]<- NA
Panel_Data["PreLevelControl"] <- NA
Panel_Data["PreTrendControl"] <- NA
for(i in 1:length(Panel_Data[[1]]))
{
Panel_Data["DecayAddControl"][i,] <- calc_add(Panel_Data, max(Panel_Data["Year"][[1]]), Panel_Data[i,]["ID"][[1]], "DecayYr")
Panel_Data["DecayAddControl100"][i,] <- calc_add(Panel_Data, max(Panel_Data["Year"][[1]]), Panel_Data[i,]["ID"][[1]], "DecayYr100")
Panel_Data["PreLevelControl"][i,] <- AOI_cells@data$lnyx_1999e[AOI_cells@data$ID == Panel_Data[i,]["ID"][[1]]]
Panel_Data["PreTrendControl"][i,] <- pre_trend_func(AOI_cells@data, Panel_Data[i,]["ID"][[1]])
}
write.csv(Panel_Data,"/Users/rbtrichler/Box Sync/MacArthur/tanzania_soc_oct2017.csv")
#write.csv(Panel_Data,"/home/aiddata/Desktop/Github/MacArthur/modelData/tanzania_soc_ONLY_JUNE.csv")
#write.csv(Panel_Data, "/home/aiddata/Desktop/Github/MacArthur/modelData/tanzania_soc.csv")
#
#
#
#
#-------------------------------------------------
#Add in additional data directly into panel dataset from AFRcells#
#Population, Baseline Protected Areas, Nighttime Lights#
#-------------------------------------------------
##Add in baseline protected area data, for pre-2001 and for pre-2008 since no projects in Tanzania until 2008
#grid covers multiple countries, but will trim down using our dataset later
pa_2000<-read.csv(paste(active_dir_path,"/ProtectedAreas_Data/merge_africa_grid_pre2001.csv",sep=""))
pa_2007<-read.csv(paste(active_dir_path,"/ProtectedAreas_Data/merge_africa_grid_wdpa_pre2008.csv",sep=""))
#pa_2000 <- read.csv("/home/aiddata/Desktop/Github/MacArthur/ProtectedAreas_Data/merge_africa_grid_pre2001.csv")
#pa_2007 <- read.csv("/home/aiddata/Desktop/Github/MacArthur/ProtectedAreas_Data/merge_africa_grid_wdpa_pre2008.csv")
#Create new column with percentage of cell covered by protected area
pa_2000$wdpapct_2000 <- NA
pa_2000$wdpapct_2000 <- pa_2000$wdpa_pre2001_africa.na.sum/pa_2000$wdpa_pre2001_africa.na.count
pa_2007$wdpapct_2007 <- NA
pa_2007$wdpapct_2007 <- pa_2007$wdpa_pre2008_africa.na.sum/pa_2007$wdpa_pre2008_africa.na.count
#Merge percentage of cell covered by protected area into AFR cell dataset for pre2001 and pre2008, drop out the sum and count columns
AFRcells<-merge(pa_2000,pa_2007,by.x="ID",by.y="ID")
AFRcells <- AFRcells[,-grep("(africa)", names(AFRcells))]
#merge AFRcells into Panel_Data
Panel_Data_add <- Panel_Data
Panel_Data_add<-merge(Panel_Data_add,AFRcells, by.x="ID",by.y="ID")
## Add in GPW4 Pop Density Data, updated data for 2000,2005,2010,2015
pop <- read.csv(paste(active_dir_path,"/GPW4_Extracts/merge_africa_grid.csv",sep=""))
Panel_Data_add<-merge(Panel_Data_add,pop,by.x="ID",by.y="ID")
#Apply 2000 values to years 2001-2004 of Panel_Data, 2005 values to years 2005-2009, 2010 values to years 2010-2014
#create it in panel form with variable name Pop
Panel_Data_add$Pop<-NA
Panel_Data_add$Pop[Panel_Data_add$Year<=2014]<-Panel_Data_add$gpw_v4_density.2010.mean[Panel_Data_add$Year<=2014]
Panel_Data_add$Pop[Panel_Data_add$Year<=2009]<-Panel_Data_add$gpw_v4_density.2005.mean[Panel_Data_add$Year<=2009]
Panel_Data_add$Pop[Panel_Data_add$Year<=2004]<-Panel_Data_add$gpw_v4_density.2000.mean[Panel_Data_add$Year<=2004]
#Maintain Pop_2000 for baseline value
names(Panel_Data_add)[names(Panel_Data_add)=="gpw_v4_density.2000.mean"]="Pop_2000"
##merge in updated dmsp ntl that matches 2001-2014 years of Panel_Data to merge into main dataset
#original ntl data that was merged in was incorrect, new ntl data merged in oct 2017
#read in nighttime lights data and rename columns
source("SciClone_functions.R")
ntl<- read.csv(paste(active_dir_path,"/ntl_extracts/merge_africa_grid2.csv",sep=""))
colnames(ntl)<-sub("v4composites_calibrated_201709.","ntl_",colnames(ntl))
colnames(ntl)<-gsub(".mean","",colnames(ntl))
#create five year trend from 2009-2013 to impute 2014 value (to match years 2001-2014)
#uses timeRangeTrend from SCI which doesn't work for data.frame format
AOI_cells_ntl<-AOI_cells[1]
AOI_cells_ntl<-merge(AOI_cells_ntl,ntl,by.x="ID",by.y="ID")
AOI_cells_ntl$ntltrend_0913<-timeRangeTrend(AOI_cells_ntl,"ntl_[0-9][0-9][0-9][0-9]",2009,2013,"ID")
AOI_cells_ntl@data$ntl_2014<-NA
AOI_cells_ntl@data$ntl_2014=AOI_cells_ntl@data$ntl_2013+AOI_cells_ntl@data$ntltrend_0913
AOI_cells_ntl@data$neg2014[AOI_cells_ntl@data$ntl_2014<0]<-1
AOI_cells_ntl@data$ntl_2014[AOI_cells_ntl@data$neg2014==1]<-0
#create ntl pre-trend for 1992-2003
AOI_cells_ntl$ntl_pretrend<-timeRangeTrend(AOI_cells_ntl,"ntl_[0-9][0-9][0-9][0-9]",1992,2007,"ID")
#create non-shape file and rename to something obvious
ntl<-AOI_cells_ntl@data
# for (i in 2:length(ntl)) {
#
# if (substr(colnames(ntl)[i], 1, 4) == "n"){
#
# name = "NTL"
# year = substr(colnames(ntl)[i], 6, 9)
# dt = paste(name,"_",year,sep="")
# colnames(ntl)[i] <- dt
# }
# }
#convert it from wide to long form and reshape into panel
#ntl_late<-ntl[c(1,82:105,107)]
ntl_long <- ntl[c(1,11:23,25)]
ntl_reshape <- grep("^ntl_",names(ntl_long))
ntl_panel <- reshape(ntl_long, varying=ntl_reshape,direction="long", idvar="ID", sep="_", timevar="Year")
ntl_panel$NTL_2007<-NA
ntl_panel$NTL_2007<-ntl_panel$ntl[ntl_panel$Year==2007]
#rename ntl column to NTL to match var names in the analysis script
colnames(ntl_panel)<-sub("ntl","NTL",colnames(ntl_panel))
#add pre-trends back into ntl_panel before merge with larger panel dataset
ntl_trends<- ntl[c(1,24,27)]
ntl_panel1<-merge(ntl_panel,ntl_trends)
ntl_panel<-ntl_panel1
#merge into Panel_Data_add
Panel_Data_add1<-merge(Panel_Data_add,ntl_panel,by=c("ID","Year"))
Panel_Data_add<-Panel_Data_add1
#write.csv(Panel_Data, "/home/aiddata/Desktop/Github/MacArthur/modelData/tanzania_infra.csv")
#Panel_Data<-read.csv("/home/aiddata/Desktop/Github/MacArthur/modelData/tanzania_infra.csv")
#write.csv(Panel_Data_add,"/home/aiddata/Desktop/Github/MacArthur/modelData/tanzania_infra_panel_data_add_AUG.csv")
write.csv(Panel_Data_add,"/Users/rbtrichler/Box Sync/MacArthur/testdata/tanzania_soc_oct2017.csv")