-
Notifications
You must be signed in to change notification settings - Fork 1
/
CLIC_Brazil_Rt_regression_analysis.R
630 lines (414 loc) · 22.4 KB
/
CLIC_Brazil_Rt_regression_analysis.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
623
624
625
626
627
628
629
630
####
### Script to run Time to event regression analyses
####
rm(list=ls())
# Main packages
library("tidyverse")
library("dplyr")
library("ggplot2")
library("MASS")
library("lmtest")
library("broom")
library("emmeans")
library("reshape")
library("plot3D")
library("gridExtra")
library("finalfit")
library("writexl")
library("tidyr")
library("multcomp")
library("Epi")
# Workflow:
rm(list=ls())
##############
### Directory set up
### Update this with your local directories
##############
dir_scripts <- "C:/github/clic_brazil/"
# set up script directories
source (paste0(dir_scripts,"CLIC_Brazil_Script_directories.R"))
# loads functions to be used for standardisation
source(paste0(dir_scripts,"CLIC_Brazil_standardisation_functions.R"))
### Functions - Source multivar functions
source(paste0(dir_scripts,"CLIC_Brazil_multivar_functions.R"))
### Load Rt prediction data
#rt_all_dat <- readRDS("CC_data/City_Case_data/Brazil/Brazil_formatted/Rt_Data/Brazil_rt_prediction-current.RDS")
#### Load data update
rt_all_dat <- readRDS(paste0(dir_Rt_data,"Brazil_rt_prediction-current-paper_v2.RDS"))
# Sample 10 places at random
# set seed so I can repeat this
# set.seed(99)
# # arbitrary number to sample
# n <- 10
# # equal weighting to cities and with no replacement
# smpl <- sample(unique(rt_all_dat$city_state), n)
# rt_all_dat_subset <- rt_all_dat[rt_all_dat$city_state %in% smpl, ]
### Drop zero values
# ggplot(data=rt_all_dat_subset2,aes(x = Date, y = Rtotalhat, color=city_state, group=city_state))+
# geom_line() +
# xlab("Date") +
# ylab("Rt") +
# scale_x_date(date_breaks = "1 week", date_labels="%d-%b") +
# #coord_cartesian(ylim = c(0.0, 3.0))
#ggsave("PM_test_results/Rt_GT0_random_10_all.png", width=40, height=16, units="cm")
### Need to define start date for analysis as local arrival = 1 case per 10000
### Merge with standardised case data
### Fetch latest data set
# update for revised analysis for the paper
load(paste0(dir_data_objects,"Brazil_BigStandard_results_16_07_21.RData"))
std_case_dat <- BigStandard$standardised_incidence
## Keep place date and standardised incidence
tmp_case_dat <- std_case_dat[c("Area", "date_end", "standardised_cases")]
## Identify date where each place reaches 1 case per 10000 (0.1 cases per 1000) )
## Keep row for first day passing threshold
start_case_dat <- tmp_case_dat %>%
group_by(Area) %>%
dplyr::filter(standardised_cases > 0.1) %>%
slice(1)
## merge to rt_all_dat on date and place
names(start_case_dat)[2] <- "Start_Date"
start_case_dat <- start_case_dat[c("Area", "Start_Date")]
rt_merge_dat <- merge(rt_all_dat,start_case_dat,by.x=c("city_state"),by.y=c("Area"),all.x=TRUE)
rt_merge_dat$day_number <- as.integer(rt_merge_dat$Date - rt_merge_dat$Start_Date)
# tmp2_case_dat <- rt_merge_dat %>%
# group_by(city_state) %>%
# arrange(city_state,Date) %>%
# slice(1)
#
# #### Histogram of max days to start
#
# ggplot(tmp2_case_dat, aes(x=day_number)) +
# geom_histogram()
#
# summary(tmp2_case_dat$day_number)
### Drop Rt predictions of 0 predictions for R0 as artefacts
rt_gt0_dat <- rt_merge_dat[ which(rt_merge_dat$Rt_Smooth > 0 ) ,]
### Start date = epidemic start date
tmp.dat <- rt_gt0_dat[c("city_state", "Start_Date")]
date_rt_start_dat <- tmp.dat %>%
group_by(city_state) %>%
slice(1)
### Mean Rt - over rnage day_start to day_end (days after Rt > 0 )
### All data
day_start <- 30
day_end <- 150
mean_rt_dat <- rt_gt0_dat %>%
dplyr::filter(day_number>=day_start & day_number <= day_end ) %>%
group_by(city_state) %>%
summarise_at(vars( "Rt_Smooth"), mean)
summary(mean_rt_dat$Rt_Smooth)
rt_mean_dat <- merge(date_rt_start_dat,mean_rt_dat,by="city_state")
names(rt_mean_dat)[3] <- "Rt_mean"
### Get day of year as a number
rt_mean_dat$start_date_day <- lubridate::yday(rt_mean_dat$Start_Date)
saveRDS(rt_mean_dat,file=(paste0(dir_Rt_data,"Brazil_mean_Rt.RDS")))
### Merge with other variables
### Covariate data created in PM_multivar_anal_v3.R
covar_dat <- readRDS(paste0(dir_Rt_data,"Brazil_lm_covariates_fail_10.RDS"))
rt_mean_covar_dat <- merge(rt_mean_dat,covar_dat,by.x="city_state",by.y="Area_char",all.x=TRUE)
## geo region as a factor
rt_mean_covar_dat$geo_region_factor <- as.factor(rt_mean_covar_dat$geo_region)
### Histogram of Rt_mean
p <- rt_mean_covar_dat %>%
ggplot( aes(x=Rt_mean)) +
geom_histogram( binwidth=0.05, fill="#69b3a2", color="#e9ecef", alpha=0.9) +
ggtitle("Rt_mean - bin = 0.05") +
theme(
plot.title = element_text(size=20)
)
p
# ggsave("PM_test_results/RT_historgram.png",p, width=20, height=15, units="cm")
rt_mean_covar_dat$log_Rt_mean <- log(rt_mean_covar_dat$Rt_mean)
### Histogram of log Rt_mean
p <- rt_mean_covar_dat %>%
ggplot( aes(x=log_Rt_mean)) +
geom_histogram( binwidth=0.05, fill="#69b3a2", color="#e9ecef", alpha=0.9) +
ggtitle("log Rt_mean - bin = 0.05") +
theme(
plot.title = element_text(size=20)
)
p
# ggsave("PM_test_results/log_RT_mean_historgram.png",p, width=20, height=15, units="cm")
## Histogram of start date day
p <- rt_mean_covar_dat %>%
ggplot( aes(x=start_date_day)) +
geom_histogram( binwidth=5, fill="#69b3a2", color="#e9ecef", alpha=0.9) +
ggtitle("Start Date Day - bin = 5") +
theme(
plot.title = element_text(size=20)
)
p
# ggsave("PM_test_results/Mid_Day_historgram.png",p, width=20, height=15, units="cm")
### Categorical variable for start date day
rt_mean_covar_dat$start_day_group <- cut(
rt_mean_covar_dat$start_date_day,
breaks = quantile(rt_mean_covar_dat$start_date_day, c(0, 0.33, 0.66, 1),na.rm = TRUE),
labels = c("gp1", "gp2", "gp3"),
right = FALSE,
include.lowest = TRUE
)
# Sumary stats
tapply(rt_mean_covar_dat$Start_Date, rt_mean_covar_dat$start_day_group, summary)
gp1 <- dplyr::filter(rt_mean_covar_dat, start_day_group=="gp1" )
min_gp1 <- format(min(gp1$Start_Date),"%d-%b-%y")
max_gp1 <- format(max(gp1$Start_Date),"%d-%b-%y")
range_gp1 <- paste0(min_gp1, " to ", max_gp1)
gp2 <- dplyr::filter(rt_mean_covar_dat, start_day_group=="gp2" )
min_gp2 <- format(min(gp2$Start_Date),"%d-%b-%y")
max_gp2 <- format(max(gp2$Start_Date),"%d-%b-%y")
range_gp2 <- paste0(min_gp2, " to ", max_gp2)
gp3<- dplyr::filter(rt_mean_covar_dat, start_day_group=="gp3" )
min_gp3 <- format(min(gp3$Start_Date),"%d-%b-%y")
max_gp3 <- format(max(gp3$Start_Date),"%d-%b-%y")
range_gp3 <- paste0(min_gp3, " to ", max_gp3)
rt_mean_covar_dat$start_day_group <- factor(rt_mean_covar_dat$start_day_group,
levels = c("gp1", "gp2" , "gp3"),
labels = c(range_gp1, range_gp2, range_gp3))
table(rt_mean_covar_dat$start_day_group)
covar_cont <- c( "log_popden","Piped_water_percent",
"Sewage_or_septic_percent", "log_travel_time_hours", "SDI_index")
## Drop rows with NA for any continuous covariate
rt_omit_dat <- na.omit(rt_mean_covar_dat, cols=covar_cont)
### Model building 2 - Start with geo region , day and geo_region*day and add other variables
covid.mod1 <- lm(Rt_mean ~ geo_region_factor*start_day_group , data = rt_omit_dat)
summary(covid.mod1)
Epi::ci.lin(covid.mod1)
### interpreting this model
#emmeans::emtrends(covid.mod1, ~ start_day_group)
covid.mod1b <- lm(Rt_mean ~ start_day_group:geo_region_factor + start_day_group + geo_region_factor , data = rt_omit_dat)
summary(covid.mod1b)
Epi::ci.lin(covid.mod1b)
## ggplot stacked bar chart of coefficient values
#https://www.r-graph-gallery.com/48-grouped-barplot-with-ggplot2.html
## useful article on intepreting models with interaction terms
# https://www.andrew.cmu.edu/user/achoulde/94842/lectures/lecture10/lecture10-94842.html
### plotting this (line plot)
emmeans::emmip(covid.mod1, start_day_group ~ geo_region_factor)
### plotting this (bar plot)
# geo_day_dat1 <- as.data.frame(emmeans::emmeans(covid.mod1, ~ start_day_group*geo_region_factor))
#
#
# p <- ggplot(data= geo_day_dat1, aes(x=geo_region_factor,y=emmean, fill=start_day_group)) +
# geom_bar(stat="identity",position="dodge") +
# geom_errorbar(position=position_dodge(.9),width=.25, aes(ymax=upper.CL, ymin=lower.CL),alpha=0.3) +
# geom_text(aes(label=sprintf("%0.2f", round(emmean, digits = 2))), position=position_dodge(width=0.9), vjust=-0.5,size=3.5) +
# labs(x="Geographic region", y="Rt Mean", fill="Mid day",
# title = "Rt Mean predictions for combinations of Mid day and Region
# \n (Unadjusted plot)") +
# coord_cartesian(ylim = c(0.9, 1.5) )
# p
# ggsave("PM_test_results/lm_Rtmean_geo_bar.png",p, width=20, height=15, units="cm")
covar_all <- c( "start_day_group" ,"geo_region_factor", "log_popden","Piped_water_percent",
"Sewage_or_septic_percent", "log_travel_time_hours", "SDI_index")
### Getting tabular output
### Get summary tabl,
summary.dat <- summ_tab(rt_omit_dat,covar_cont,2)
### add start point day
tmp.dat <- as.data.frame(table(rt_omit_dat$start_day_group))
names(tmp.dat)[1] <- "term"
names(tmp.dat)[2] <- "dist"
tmp.dat$term <- paste0("start_day_group",tmp.dat$term)
summary.dat <- rbind(tmp.dat,summary.dat)
# ### add geo region
tmp.dat <- as.data.frame(table(rt_omit_dat$geo_region))
names(tmp.dat)[1] <- "term"
names(tmp.dat)[2] <- "dist"
tmp.dat$term <- paste0("geo_region_factor",tmp.dat$term)
summary.dat <- rbind(tmp.dat,summary.dat)
## keep row numbers
summary.dat <- as.data.frame(data.table::setDT(summary.dat, keep.rownames = TRUE)[])
summary.dat$rn <- as.numeric(summary.dat$rn)
### Summarise by group
# rt_omit_dat %>%
# group_by(geo_region_factor) %>%
# summarise(mean(Rt_mean))
#
# levels(rt_omit_dat$start_day_group)
### Get univariate stats
univar.dat <- uni_tab(rt_omit_dat,"Rt_mean ~ ",covar_all,2,3)
# covar_cont <- c( "log_popden","Piped_water_percent","Sewage_or_septic_percent", "log_travel_time_hours", "SDI")
## Suggests that we should drop GDP variable
covid.mod3 <- lm(Rt_mean ~ start_day_group*geo_region_factor + log_popden , data = rt_omit_dat)
anova(covid.mod3,covid.mod1)
## keep log popden
covid.mod4 <- lm(Rt_mean ~ start_day_group*geo_region_factor + log_popden + Piped_water_percent , data = rt_omit_dat)
anova(covid.mod3,covid.mod4)
## keep piped water
covid.mod5 <- lm(Rt_mean ~ start_day_group*geo_region_factor + log_popden + Piped_water_percent + Sewage_or_septic_percent , data = rt_omit_dat)
anova(covid.mod5,covid.mod4)
## keep Sewage
covid.mod6 <- lm(Rt_mean ~ start_day_group*geo_region_factor + log_popden + Piped_water_percent + Sewage_or_septic_percent
+ log_travel_time_hours , data = rt_omit_dat)
anova(covid.mod6,covid.mod5)
## keep travel
covid.mod7 <- lm(Rt_mean ~ start_day_group*geo_region_factor + log_popden + Piped_water_percent + Sewage_or_septic_percent
+ log_travel_time_hours + SDI_index , data = rt_omit_dat)
anova(covid.mod7,covid.mod6)
# Keep SDI
### Multivariate output
multivar.dat <- multi_tab(rt_omit_dat,"Rt_mean ~ start_day_group*geo_region_factor + log_popden + Piped_water_percent + Sewage_or_septic_percent + log_travel_time_hours + SDI_index",2,3)
## fix column numbers ###
final_table.dat <- merge_sum_uni_mult(summary.dat,univar.dat,multivar.dat)
## Fix p value estimates
final_table.dat$Univariate_p_value <- ifelse(final_table.dat$Univariate_p_value=="0.00", "<0.01", final_table.dat$Univariate_p_value)
final_table.dat$Multivariate_p_value <- ifelse(final_table.dat$Multivariate_p_value=="0.00", "<0.01", final_table.dat$Multivariate_p_value)
writexl::write_xlsx(final_table.dat,paste0(dir_results,"RT_regress_final_table_",as.character(day_start),"_",as.character(day_end),".xlsx"))
geo_day_dat1 <- as.data.frame(emmeans::emmeans(covid.mod7, ~ start_day_group*geo_region_factor))
geo_day_dat1$start_day_group <- factor(geo_day_dat1$start_day_group,
levels = c("gp1", "gp2" , "gp3"),
labels = c(range_gp1, range_gp2, range_gp3))
p <- ggplot(data= geo_day_dat1, aes(x=geo_region_factor,y=emmean, fill=start_day_group)) +
geom_bar(stat="identity",position="dodge") +
geom_errorbar(position=position_dodge(.9),width=.25, aes(ymax=upper.CL, ymin=lower.CL),alpha=0.3,colour="grey50") +
geom_text(aes(label=sprintf("%0.2f", round(emmean, digits = 2))), position=position_dodge(width=0.9), vjust=-0.5,size=3.5) +
labs(x="Geographic region", y="Rt Mean", fill="Date of local epidemic start",
title = "") +
coord_cartesian(ylim = c(0.8,1.25) )
p
#ggsave("PM_test_results/lm_Rtmean_geo_bar.png",p, width=20, height=15, units="cm")
### Export the table
summary(rt_mean_covar_dat$Rt_mean)
### Plotting estimates for adjusted model
inter_file <- paste0("PM_test_results/RT_regress_interact_estimates_",as.character(day_start),"_",as.character(day_end),".xlsx")
interact_estimates.dat <- as.data.frame(Epi::ci.lin(covid.mod7))
interact_estimates.dat$Variable <- rownames(interact_estimates.dat)
names(interact_estimates.dat)[4] <- "p_value"
names(interact_estimates.dat)[5] <- "L95CI"
names(interact_estimates.dat)[6] <- "U95CI"
interact_estimates.dat$Estimate <- formatC(interact_estimates.dat$Estimate, format = "f", digits = 3)
interact_estimates.dat$p_value <- formatC(interact_estimates.dat$p_value, format = "f", digits = 3)
interact_estimates.dat$p_value <- ifelse(interact_estimates.dat$p_value=="0.000", "<0.001", interact_estimates.dat$p_value)
interact_estimates.dat$L95CI <- formatC(interact_estimates.dat$L95CI, format = "f", digits = 3)
interact_estimates.dat$U95CI <- formatC(interact_estimates.dat$U95CI, format = "f", digits = 3)
interact_estimates.dat$est_ci <- paste0(interact_estimates.dat$Estimate," (",interact_estimates.dat$L95CI," - "
,interact_estimates.dat$U95CI,")")
#write.xlsx(interact_estimates.dat ,file = inter_file, row.names = FALSE)
#interact_estimates.dat <- interact_estimates.dat[c(7,8,4)]
## Need to figure out how to do this in code
bar_data.dat <- read.csv2(file="PM_test_results/RT_regress_interact_estimates_30_150_bar_chart.csv",sep=",")
names(bar_data.dat)[1] <- "Estimate"
bar_data.dat$Estimate <- as.numeric(bar_data.dat$Estimate )
bar_data.dat$L95CI <- as.numeric(bar_data.dat$L95CI )
bar_data.dat$U95CI <- as.numeric(bar_data.dat$U95CI )
bar_data.dat$start_day <- factor(bar_data.dat$start_day , levels = c("24-Apr to 31-May","1-Jun to 30-Jun","1-Jul to 14-Sep"))
bar_data.dat$Estimate_label <- formatC(bar_data.dat$Estimate, format = "f", digits = 2)
p <- ggplot(bar_data.dat, aes(fill=start_day, y=Estimate, x=geo_region)) +
geom_bar(position="dodge", stat="identity",colour="black",width=0.8) +
geom_errorbar(aes(ymin=L95CI, ymax=U95CI), width=.2,position=position_dodge(0.9),colour="grey50") +
geom_abline(slope=0, intercept=0.0, col = "black") +
geom_text(aes(label=Estimate_label), position=position_dodge(width=0.9), vjust=-0.8,colour="black",size = 3.0) +
labs(y = "Estimate",
x = "Geographic region",
fill="")
p
ggsave("PM_test_results/interaction_estimates.png",p, width=20, height=15, units="cm")
#
# bar_data.dat$start_day_group <- bar_data.dat$Variable
# bar_data.dat$geo_region <- ""
#
#
#
#
# ## Sort out start_day_group
# # make column blank if not start_day_group
# bar_data.dat$start_day_group[!grepl("start_day_group",bar_data.dat$start_day_group)]<-""
#
# bar_data.dat$start_day_group[bar_data.dat$start_day_group == "DEF"] <-"NEW1"
## Get data in the format for grouped bar chart
write.xlsx(interact_estimates.dat ,file = inter_file, row.names = FALSE)
## Check range of Rtmean
summary(rt_omit_dat$Rt_mean)
summary(covid.mod7)
adj_mean_rt.dat <- as.data.frame(emmeans::emmeans(covid.mod7, ~ start_day_group | geo_region_factor))
### table of frequencies
tmp.dat <- as.data.frame(table(rt_omit_dat$geo_region_factor,rt_omit_dat$start_day_group))
tmp.dat <- tmp.dat %>% dplyr::arrange(Var1,Var2)
names(tmp.dat)[1] <- "geo_region_factor"
names(tmp.dat)[2] <- "start_day_group"
# Set row name to 1st column
tmp.dat <- as.data.frame(data.table::setDT(tmp.dat, keep.rownames = TRUE)[])
tmp.dat$rn <- as.numeric(tmp.dat$rn)
Mean_Rt_summary.dat <- merge(tmp.dat,adj_mean_rt.dat,by=c("geo_region_factor","start_day_group"))
Mean_Rt_summary.dat$emmean_c <- formatC(Mean_Rt_summary.dat$emmean, format = "f", digits = 3)
Mean_Rt_summary.dat$lower.CL_c <- formatC(Mean_Rt_summary.dat$lower.CL, format = "f", digits = 3)
Mean_Rt_summary.dat$upper.CL_c <- formatC(Mean_Rt_summary.dat$upper.CL, format = "f", digits = 3)
Mean_Rt_summary.dat$est_ci <- paste0(Mean_Rt_summary.dat$emmean_c," (",Mean_Rt_summary.dat$lower.CL_c," - ",Mean_Rt_summary.dat$upper.CL_c,") ")
Mean_Rt_summary.dat <- Mean_Rt_summary.dat %>% dplyr::arrange(rn)
#Mean_Rt_summary.dat <- Mean_Rt_summary.dat[c(1,2,4,10)]
write.xlsx(Mean_Rt_summary.dat ,file = "plots/calc_Rt_geo_day.xlsx", row.names = FALSE)
## reorder levels of geo region
Mean_Rt_summary.dat$geo_region_factor <- factor(Mean_Rt_summary.dat$geo_region_factor, levels = c("N","NE","CW","SE","S") )
#levels(Mean_Rt_summary.dat$geo_region_factor)
p <- ggplot(data= Mean_Rt_summary.dat, aes(x=geo_region_factor,y=emmean, fill=start_day_group)) +
geom_bar(stat="identity",position="dodge") +
geom_errorbar(position=position_dodge(.9),width=.25, aes(ymax=upper.CL, ymin=lower.CL),alpha=0.3) +
geom_text(aes(label=sprintf("%0.2f", round(emmean, digits = 2))), position=position_dodge(width=0.9), vjust=-0.5,size=3.5) +
labs(x="Geographic region", y="Rt Mean", fill="Mid day",
title = "Rt Mean predictions for combinations of Mid day and Region (Rt mean window - 30 - 100 days)
\n (Adjusted model multivariate model )") +
coord_cartesian(ylim = c(0.8, 1.2) )
p
ggsave("PM_test_results/lm_Rtmean_geo_adj_bar-30-100.png",p, width=30, height=15, units="cm")
### Investigation of the use of a log Rt mean in the output
rt_omit_dat$log_Rt_mean <- log(rt_omit_dat$Rt_mean)
### Get univariate stats
univar_logrt.dat <- uni_tab(rt_omit_dat,"log_Rt_mean ~ ",covar_all,2,3)
### Multivariate output
multivar_logrt.dat <- multi_tab(rt_omit_dat,"log_Rt_mean ~ start_day_group*geo_region_factor + log_popden + Piped_water_percent + Sewage_or_septic_percent + log_travel_time_hours + SDI_index",2,3)
## transform estimates from the log scale
#formatC(interact_estimates.dat$L95CI, format = "f", digits = 3)
univar_logrt.dat$estimate <- as.character(formatC(exp(as.numeric(univar_logrt.dat$estimate))), format = "f", digits = 3)
univar_logrt.dat$CI_2.5 <- as.character(formatC(exp(as.numeric(univar_logrt.dat$CI_2.5))), format = "f", digits = 3)
univar_logrt.dat$CI_97.5 <- as.character(formatC(exp(as.numeric(univar_logrt.dat$CI_97.5))), format = "f", digits = 3)
multivar_logrt.dat$estimate <- as.character(formatC(exp(as.numeric(multivar_logrt.dat$estimate))), format = "f", digits = 3)
multivar_logrt.dat$CI_2.5 <- as.character(formatC(exp(as.numeric(multivar_logrt.dat$CI_2.5))), format = "f", digits = 3)
multivar_logrt.dat$CI_97.5 <- as.character(formatC(exp(as.numeric(multivar_logrt.dat$CI_97.5))), format = "f", digits = 3)
final_table_logrt.dat <- merge_sum_uni_mult(summary.dat,univar_logrt.dat,multivar_logrt.dat)
## Fix p value estimates
final_table_logrt.dat$Univariate_p_value <- ifelse(final_table_logrt.dat$Univariate_p_value=="0.00", "<0.01", final_table_logrt.dat$Univariate_p_value)
final_table_logrt.dat$Multivariate_p_value <- ifelse(final_table_logrt.dat$Multivariate_p_value=="0.00", "<0.01", final_table_logrt.dat$Multivariate_p_value)
writexl::write_xlsx(final_table_logrt.dat,paste0(dir_results,"RT_regress_logrt_",as.character(day_start),"_",as.character(day_end),".xlsx"))
### Transforning interaction terms
covid.mod8 <- lm(log_Rt_mean ~ start_day_group*geo_region_factor + log_popden + Piped_water_percent + Sewage_or_septic_percent
+ log_travel_time_hours + SDI_index , data = rt_omit_dat)
adj_mean_rt.dat <- as.data.frame(emmeans::emmeans(covid.mod8, ~ start_day_group | geo_region_factor))
### exponentiate estimates
adj_mean_rt.dat$emmean_c <- as.character(formatC(exp(as.numeric(adj_mean_rt.dat$emmean))), format = "f", digits = 3)
adj_mean_rt.dat$lower.CL_c <- as.character(formatC(exp(as.numeric(adj_mean_rt.dat$lower.CL))), format = "f", digits = 3)
adj_mean_rt.dat$upper.CL_c <- as.character(formatC(exp(as.numeric(adj_mean_rt.dat$upper.CL))), format = "f", digits = 3)
adj_mean_rt.dat$emmean <- exp(as.numeric(adj_mean_rt.dat$emmean))
adj_mean_rt.dat$lower.CL <- exp(as.numeric(adj_mean_rt.dat$lower.CL))
adj_mean_rt.dat$upper.CL <- exp(as.numeric(adj_mean_rt.dat$upper.CL))
### table of frequencies
tmp.dat <- as.data.frame(table(rt_omit_dat$geo_region_factor,rt_omit_dat$start_day_group))
tmp.dat <- tmp.dat %>% dplyr::arrange(Var1,Var2)
names(tmp.dat)[1] <- "geo_region_factor"
names(tmp.dat)[2] <- "start_day_group"
# Set row name to 1st column
tmp.dat <- as.data.frame(data.table::setDT(tmp.dat, keep.rownames = TRUE)[])
tmp.dat$rn <- as.numeric(tmp.dat$rn)
log_Mean_Rt_summary.dat <- merge(tmp.dat,adj_mean_rt.dat,by=c("geo_region_factor","start_day_group"))
log_Mean_Rt_summary.dat$est_ci <- paste0(log_Mean_Rt_summary.dat$emmean_c," (",log_Mean_Rt_summary.dat$lower.CL_c," - ",log_Mean_Rt_summary.dat$upper.CL_c,") ")
log_Mean_Rt_summary.dat <- log_Mean_Rt_summary.dat %>% dplyr::arrange(rn)
#Mean_Rt_summary.dat <- Mean_Rt_summary.dat[c(1,2,4,10)]
#write.xlsx(log_Mean_Rt_summary.dat ,file = "plots/calc_log_Rt_geo_day.xlsx", row.names = FALSE)
#
## reorder levels of geo region
log_Mean_Rt_summary.dat$geo_region_factor <- factor(log_Mean_Rt_summary.dat$geo_region_factor, levels = c("N","NE","CW","SE","S") )
#levels(Mean_Rt_summary.dat$geo_region_factor)
p <- ggplot(data= log_Mean_Rt_summary.dat, aes(x=geo_region_factor,y=emmean, fill=start_day_group)) +
geom_bar(stat="identity",position="dodge") +
geom_errorbar(position=position_dodge(.9),width=.25, aes(ymax=upper.CL, ymin=lower.CL),alpha=0.3) +
geom_text(aes(label=sprintf("%0.2f", round(emmean, digits = 2))), position=position_dodge(width=0.9), vjust=-0.5,size=3.5) +
labs( fill="Range of start dates for calculation of mean", title = "") +
xlab("Geographic region") +
ylab(expression("Mean of "*italic(R[t])*"")) +
coord_cartesian(ylim = c(0.50, 1.3) )
p
ggsave( paste0(dir_results,"lm_log_Rtmean_geo_adj_bar-30-150.png"),p, width=30, height=15, units="cm")
# ggplot(data= log_Mean_Rt_summary.dat, aes(x=geo_region_factor,y=emmean, fill=start_day_group)) +
# geom_bar(stat="identity",position="dodge")
## testing
# lm(Rt_mean ~ log_popden , data = rt_omit_dat)
#
# lm(log_Rt_mean ~ log_popden , data = rt_omit_dat)