Merge pull request #4 from cmu-delphi/ndefries/match-scripts-to-epiprocess

Move data script formatting improvements from `epiprocess`

Author: nmdefries

DESCRIPTION

@@ -28,5 +28,5 @@ Remotes:
     cmu-delphi/epiprocess
 Encoding: UTF-8
 LazyData: true
-RoxygenNote: 7.2.3
+RoxygenNote: 7.3.1
 URL: https://cmu-delphi.github.io/epidatasets/

data-raw/archive_cases_dv_subset_dt.R

@@ -1,14 +1,15 @@
-library(dplyr)
 library(epidatr)
 library(epiprocess)
+library(data.table)
+library(dplyr)
 
 dv_subset <- pub_covidcast(
   source = "doctor-visits",
   signals = "smoothed_adj_cli",
-  time_type = "day",
   geo_type = "state",
-  time_values = epirange(20200601, 20211201),
+  time_type = "day",
   geo_values = "ca,fl,ny,tx",
+  time_values = epirange(20200601, 20211201),
   issues = epirange(20200601, 20211201)
 ) %>%
   select(geo_value, time_value, version = issue, percent_cli = value) %>%
@@ -19,16 +20,16 @@ dv_subset <- pub_covidcast(
 case_rate_subset <- pub_covidcast(
   source = "jhu-csse",
   signals = "confirmed_7dav_incidence_prop",
-  time_type = "day",
   geo_type = "state",
-  time_values = epirange(20200601, 20211201),
+  time_type = "day",
   geo_values = "ca,fl,ny,tx",
+  time_values = epirange(20200601, 20211201),
   issues = epirange(20200601, 20211201)
 ) %>%
   select(geo_value, time_value, version = issue, case_rate_7d_av = value) %>%
   as_epi_archive(compactify = FALSE)
 
-archive_cases_dv_subset = epix_merge(
+archive_cases_dv_subset <- epix_merge(
   dv_subset, case_rate_subset,
   sync = "locf",
   compactify = FALSE)
@@ -39,4 +40,4 @@ archive_cases_dv_subset = epix_merge(
 # objects; store the DT and construct the R6 object on request.
 archive_cases_dv_subset_dt <- archive_cases_dv_subset$DT
 
-usethis::use_data(archive_cases_dv_subset_dt, overwrite = TRUE)
+usethis::use_data(archive_cases_dv_subset_dt, overwrite = TRUE, internal = TRUE)

data-raw/cases_deaths_subset.R

@@ -1,58 +1,61 @@
-library(dplyr)
 library(epidatr)
 library(epiprocess)
+library(dplyr)
 
-confirmed_7dav_incidence_prop <- pub_covidcast(
+confirmed_incidence_num <- pub_covidcast(
   source = "jhu-csse",
-  signals = "confirmed_7dav_incidence_prop",
-  time_type = "day",
+  signals = "confirmed_incidence_num",
   geo_type = "state",
+  time_type = "day",
+  geo_values = "ca,fl,ny,tx,ga,pa",
   time_values = epirange(20200301, 20211231),
-  geo_values = "ca,fl,ny,tx,ga,pa"
 ) %>%
-  select(geo_value, time_value, case_rate_7d_av = value) %>%
+  select(geo_value, time_value, cases = value) %>%
   arrange(geo_value, time_value)
 
-deaths_7dav_incidence_prop <- pub_covidcast(
+confirmed_7dav_incidence_num <- pub_covidcast(
   source = "jhu-csse",
-  signals = "deaths_7dav_incidence_prop",
-  time_type = "day",
+  signals = "confirmed_7dav_incidence_num",
   geo_type = "state",
+  time_type = "day",
+  geo_values = "ca,fl,ny,tx,ga,pa",
   time_values = epirange(20200301, 20211231),
-  geo_values = "ca,fl,ny,tx,ga,pa"
 ) %>%
-  select(geo_value, time_value, death_rate_7d_av = value) %>%
+  select(geo_value, time_value, cases_7d_av = value) %>%
   arrange(geo_value, time_value)
 
-confirmed_incidence_num <- pub_covidcast(
+confirmed_7dav_incidence_prop <- pub_covidcast(
   source = "jhu-csse",
-  signals = "confirmed_incidence_num",
-  time_type = "day",
+  signals = "confirmed_7dav_incidence_prop",
   geo_type = "state",
+  time_type = "day",
+  geo_values = "ca,fl,ny,tx,ga,pa",
   time_values = epirange(20200301, 20211231),
-  geo_values = "ca,fl,ny,tx,ga,pa"
 ) %>%
-  select(geo_value, time_value, cases = value) %>%
+  select(geo_value, time_value, case_rate_7d_av = value) %>%
   arrange(geo_value, time_value)
 
-confirmed_7dav_incidence_num <- pub_covidcast(
+deaths_7dav_incidence_prop <- pub_covidcast(
   source = "jhu-csse",
-  signals = "confirmed_7dav_incidence_num",
-  time_type = "day",
+  signals = "deaths_7dav_incidence_prop",
   geo_type = "state",
+  time_type = "day",
+  geo_values = "ca,fl,ny,tx,ga,pa",
   time_values = epirange(20200301, 20211231),
-  geo_values = "ca,fl,ny,tx,ga,pa"
 ) %>%
-  select(geo_value, time_value, cases_7d_av = value) %>%
+  select(geo_value, time_value, death_rate_7d_av = value) %>%
   arrange(geo_value, time_value)
 
-cases_deaths_subset <- confirmed_7dav_incidence_prop %>%
-  full_join(deaths_7dav_incidence_prop,
-            by = c("geo_value", "time_value")) %>%
-  full_join(confirmed_incidence_num,
-            by = c("geo_value", "time_value")) %>%
+cases_deaths_subset <- confirmed_incidence_num %>%
   full_join(confirmed_7dav_incidence_num,
-            by = c("geo_value", "time_value")) %>%
+    by = c("geo_value", "time_value")
+  ) %>%
+  full_join(confirmed_7dav_incidence_prop,
+    by = c("geo_value", "time_value")
+  ) %>%
+  full_join(deaths_7dav_incidence_prop,
+    by = c("geo_value", "time_value")
+  ) %>%
   as_epi_df()
 
 usethis::use_data(cases_deaths_subset, overwrite = TRUE)

data-raw/covid_incidence_county_subset.R

@@ -1,9 +1,9 @@
-# Use covidcast::county_census to get the county and state names
-library(dplyr)
-library(covidcast)
 library(epidatr)
+library(covidcast)
 library(epiprocess)
+library(dplyr)
 
+# Use covidcast::county_census to get the county and state names
 y <- covidcast::county_census %>%
   filter(STNAME %in% c("Massachusetts", "Vermont"), STNAME != CTYNAME) %>%
   select(geo_value = FIPS, county_name = CTYNAME, state_name = STNAME)
@@ -12,10 +12,10 @@ y <- covidcast::county_census %>%
 covid_incidence_county_subset <- pub_covidcast(
   source = "jhu-csse",
   signals = "confirmed_incidence_num",
-  time_type = "day",
   geo_type = "county",
+  time_type = "day",
+  geo_values = paste(y$geo_value, collapse = ","),
   time_values = epirange(20200601, 20211231),
-  geo_values = paste(y$geo_value, collapse = ",")
 ) %>%
   select(geo_value, time_value, cases = value) %>%
   full_join(y, by = "geo_value") %>%

data-raw/covid_incidence_outliers.R

@@ -1,14 +1,15 @@
-library(dplyr)
 library(epidatr)
 library(epiprocess)
+library(dplyr)
+library(tidyr)
 
 covid_incidence_outliers <- pub_covidcast(
   source = "jhu-csse",
   signals = "confirmed_incidence_num",
-  time_type = "day",
   geo_type = "state",
-  time_values = epirange(20200601, 20210531),
+  time_type = "day",
   geo_values = "fl,nj",
+  time_values = epirange(20200601, 20210531),
   as_of = 20211028
 ) %>%
   select(geo_value, time_value, cases = value) %>%