Merge pull request #36 from RhoInc/dashboard-data-updates

Update enrollment data

.gitignore

@@ -1,3 +1,5 @@
 .DS_Store
 
 node_modules/
+
+.Rout

data/clinical-trials/adam/scripts/run-scripts.bat

@@ -1,10 +1,7 @@
 rem Make sure Rpath is pointing to your local install of R.
-    set Rpath="C:\Program Files\R\R-3.3.2\bin\R.exe"
-        %Rpath% CMD BATCH --no-save --vanilla --slave --quiet create_DM.R
-        %Rpath% CMD BATCH --no-save --vanilla --slave --quiet create_AE.R
-        %Rpath% CMD BATCH --no-save --vanilla --slave --quiet create_CM.R
-        %Rpath% CMD BATCH --no-save --vanilla --slave --quiet create_SV.R
-        %Rpath% CMD BATCH --no-save --vanilla --slave --quiet labs.R
-        %Rpath% CMD BATCH --no-save --vanilla --slave --quiet create_LB.R
-        %Rpath% CMD BATCH --no-save --vanilla --slave --quiet create_VS.R
-        %Rpath% CMD BATCH --no-save --vanilla --slave --quiet create_anly.R
+    set Rpath="C:\Program Files\R\R-3.4.3\bin\R.exe"
+        %Rpath% CMD BATCH --no-save --vanilla --slave --quiet adae.R
+        %Rpath% CMD BATCH --no-save --vanilla --slave --quiet adcm.R
+        %Rpath% CMD BATCH --no-save --vanilla --slave --quiet adlb.R
+        %Rpath% CMD BATCH --no-save --vanilla --slave --quiet adsl.R
+        %Rpath% CMD BATCH --no-save --vanilla --slave --quiet advs.R

data/clinical-trials/data-cleaning/data-files.json

@@ -1,32 +1,40 @@
 [
     {
-        "relPath": "./data/clinical-trials/data-cleaning/dashboard-enrollment-over-time.csv",
-        "path": "dashboard-enrollment-over-time.csv",
-        "name": "dashboard-enrollment-over-time",
-        "nRows": 1206,
+        "relPath": "./data/clinical-trials/data-cleaning/dashboard-accrual-over-time.csv",
+        "path": "dashboard-accrual-over-time.csv",
+        "name": "dashboard-accrual-over-time",
+        "nRows": 6051,
         "cols": [
             "site",
-            "date",
             "population",
             "population_order",
             "population_color",
+            "date",
             "participant_count"
         ],
         "nCols": 6
     },
     {
-        "relPath": "./data/clinical-trials/data-cleaning/dashboard-enrollment.csv",
-        "path": "dashboard-enrollment.csv",
-        "name": "dashboard-enrollment",
-        "nRows": 69,
+        "relPath": "./data/clinical-trials/data-cleaning/dashboard-accrual.csv",
+        "path": "dashboard-accrual.csv",
+        "name": "dashboard-accrual",
+        "nRows": 263,
         "cols": [
-            "site",
+            "subjid",
             "population",
             "population_order",
             "population_color",
-            "population_superset"
+            "population_superset",
+            "date",
+            "site",
+            "site_abbreviation",
+            "site_info",
+            "filter:Arm",
+            "listing:Age",
+            "listing:Sex",
+            "listing:Race"
         ],
-        "nCols": 5
+        "nCols": 13
     },
     {
         "relPath": "./data/clinical-trials/data-cleaning/dashboard-forms.csv",
@@ -137,7 +145,6 @@
             "subjectnameoridentifier",
             "subject_status",
             "visit_name",
-            "visit_abbreviation",
             "visit_number",
             "visit_date",
             "visit_day",
@@ -152,6 +159,6 @@
             "overdue2",
             "plot_exclude"
         ],
-        "nCols": 18
+        "nCols": 17
     }
 ]

data/clinical-trials/data-cleaning/scripts/dashboard-accrual-over-time.R

@@ -0,0 +1,132 @@
+library(data.table)
+library(tidyverse)
+library(lubridate)
+
+# input data
+accrual <- '../dashboard-accrual.csv' %>%
+    fread(
+        sep = ',',
+        na.strings = '',
+        colClasses = 'character'
+    )
+sites <- '../../data-dictionaries/sites.csv' %>%
+    fread(
+        sep = ',',
+        na.strings = '',
+        colClasses = 'character'
+    )
+
+# data manipulation
+
+    # target
+    target <- sites %>%
+        mutate(
+            population = 'Target',
+            population_order = 3,
+            population_color = '#034e7b',
+            population_superset = '',
+            target_rate = as.numeric(site_target)/as.numeric(site_accrual_duration)
+        ) %>%
+        group_by(
+            population, population_order, population_color, population_superset, site, site_abbreviation, site_accrual_start_date, accrual_end_date, target_rate
+        ) %>%
+        do(data.frame(
+            date = seq(ymd(.$site_accrual_start_date), ymd(.$accrual_end_date), by = '1 day'),
+            stringsAsFactors = FALSE
+        )) %>%
+        mutate(
+            participant_count = cumsum(target_rate)
+        ) %>%
+        ungroup() %>%
+        select(
+            -site_accrual_start_date, -accrual_end_date, -target_rate
+        ) %>%
+        mutate(
+            date = as.character(date)
+        )
+
+    # screened and randomized
+    shell <- target %>%
+        select(
+            -starts_with('population'), -participant_count
+        )
+
+        # screened
+        screened <- accrual %>%
+            filter(
+                population == 'Screened'
+            ) %>%
+            select(
+                starts_with('population')
+            ) %>%
+            unique
+        screened1 <- shell %>%
+            mutate(
+                population = rep(screened$population, nrow(shell)),
+                population_order = rep(screened$population_order, nrow(shell)),
+                population_color = rep(screened$population_color, nrow(shell)),
+                population_superset = rep(screened$population_superset, nrow(shell))
+            )
+
+        # randomized
+        randomized <- accrual %>%
+            filter(
+                population == 'Randomized'
+            ) %>%
+            select(
+                starts_with('population')
+            ) %>%
+            unique
+        randomized1 <- shell %>%
+            mutate(
+                population = rep(randomized$population, nrow(shell)),
+                population_order = rep(randomized$population_order, nrow(shell)),
+                population_color = rep(randomized$population_color, nrow(shell)),
+                population_superset = rep(randomized$population_superset, nrow(shell))
+            )
+
+    # count participants by date, population, and site
+    participantCounts <- accrual %>%
+        group_by(
+            population, population_order, population_color, population_superset, date, site, site_abbreviation
+        ) %>%
+        summarize(
+            participant_count = n()
+        ) %>%
+        ungroup()
+
+    # merge population shells with participant counts
+    accrualOverTime <- screened1 %>%
+        rbind(randomized1) %>%
+        left_join(
+            participantCounts
+        ) %>%
+        mutate(
+            participant_count = ifelse(
+                is.na(participant_count),
+                0,
+                participant_count
+            )
+        ) %>%
+        group_by(
+            population, population_order, population_color, population_superset, site, site_abbreviation
+        ) %>%
+        mutate(
+            participant_count = cumsum(participant_count) # cumulatively sum participant counts by population and site
+        ) %>%
+        ungroup %>%
+        rbind(target) %>%
+        arrange(
+            site, population_order, date
+        ) %>%
+        select(
+            site, population, population_order, population_color, date, participant_count
+        )
+
+# output data
+accrualOverTime %>%
+    fwrite(
+        '../../data-cleaning/dashboard-accrual-over-time.csv',
+        na = '',
+        row.names = FALSE
+    )

data/clinical-trials/data-cleaning/scripts/dashboard-accrual.R

@@ -0,0 +1,89 @@
+library(data.table)
+library(tidyverse)
+
+# input data
+dm <- '../../sdtm/dm.csv' %>%
+    fread(
+        sep = ',',
+        na.strings = '',
+        colClasses = 'character'
+    )
+sv <- '../../sdtm/sv.csv' %>%
+    fread(
+        sep = ',',
+        na.strings = '',
+        colClasses = 'character'
+    )
+sites <- '../../data-dictionaries/sites.csv' %>%
+    fread(
+        sep = ',',
+        na.strings = '',
+        colClasses = 'character'
+    )
+
+# data manipulation
+
+    # screened population
+    screened <- dm %>%
+        select(
+            USUBJID, RFSTDTC
+        ) %>%
+        rename(
+            subjid = USUBJID,
+            date = RFSTDTC
+        ) %>%
+        mutate(
+            population = 'Screened',
+            population_order = 1,
+            population_color = '#a6bddb',
+            population_superset = ''
+        )
+
+    # randomized popluation
+    randomized <- sv %>%
+        filter(
+            VISIT == 'Visit 1' & SVSTATUS == 'Completed'
+        ) %>%
+        select(USUBJID, SVDT) %>%
+        rename(
+            subjid = USUBJID,
+            date = SVDT
+        ) %>%
+        mutate(
+            population = 'Randomized',
+            population_order = 2,
+            population_color = '#3690c0',
+            population_superset = 'Screened'
+        )
+
+    # stacked, merged with DM, merged with sites
+    accrual <- screened %>%
+        rbind(randomized) %>%
+        left_join(
+            dm %>% select(USUBJID, SITEID, AGE, SEX, RACE, ARM),
+            by = c('subjid' = 'USUBJID')
+        ) %>%
+        left_join(
+            sites %>% select(site, site_id, site_abbreviation, site_info),
+            by = c('SITEID' = 'site_id')
+        ) %>%
+        select(
+            subjid, population, population_order, population_color, population_superset, date, site, site_abbreviation, site_info, ARM, AGE, SEX, RACE
+        ) %>%
+        rename(
+            `filter:Arm` = ARM,
+            `listing:Age` = AGE,
+            `listing:Sex` = SEX,
+            `listing:Race` = RACE
+        ) %>%
+        arrange(
+            subjid, population_order
+        )
+
+# output data
+accrual %>%
+    fwrite(
+        '../../data-cleaning/dashboard-accrual.csv',
+        na = '',
+        row.names = FALSE
+    )