# Speed up model using real data
# Anna Moeller
# 7/1/2020
library(tidyverse)
# Load pictures
load("../CameraTrapStudy/2015 data/pics.wide20160804.RData")
# Make dataframe
df <- pics %>%
select(cam, timeGMT, elkpresent) %>%
filter(elkpresent == T) %>%
mutate(count = as.numeric(elkpresent)
) %>%
rename(datetime = timeGMT)
# Make a pretend deploy, as if all always working
deploy <- pics %>%
distinct(cam, op.start, op.end) %>%
# fix problem child
mutate(op.start = replace(op.start, cam == "AM158", "2016-01-05 17:00:00") ) %>%
mutate(start = as.POSIXct(op.start),
start = lubridate::force_tz(start, "GMT"),
end = as.POSIXct(op.end),
end = lubridate::force_tz(end, "GMT"),
area = 250
) %>%
select(-op.start, -op.end)
study_dates <- as.POSIXct(c("2016-01-01 01:04:18", "2016-03-27 00:00:00"),
tz = "GMT")
# Now actually use the package
occ <- build_occ(samp_freq = 40000,
samp_length = 1,
study_start = study_dates[1],
study_end = study_dates[2])
# Problem 1. speed up effort_fn.
# Problem is in group_by and summarize to get rid of multiple rows
# I want to fix the problem by getting rid of the second row.
# # play with deploy to debug
# deploy <- tibble::tibble(
# cam = c("AM174", "AM174"),
# start = as.POSIXct(c("2016-01-06 17:00:00", "2016-01-08 17:00:02"), tz = "GMT"),
# # end = "2016-01-08 17:00:01", "2016-01-10 18:00:00" ), tz = "GMT"),
# area = c(250, 100)
# )
# Test it out
ste_eh <- ste_build_eh(df, deploy, occ, quiet = T)
##################################################################################
# Time to event
tte_occ <- build_occ(samp_freq = 3600*10,
samp_length = 3600*8,
study_start = study_dates[1],
study_end = study_dates[2])
#()
effort <- effort_fn(deploy, tte_occ)
# Run it
per <- tte_samp_per(deploy, lps = 36/3600)
tte_eh <- tte_build_eh(df, deploy, tte_occ, per)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.