Add phone sensed bins (sensor row count for every n min bin)

Snakefile

@@ -11,6 +11,7 @@ rule all:
         expand("data/processed/{pid}/screen_deltas.csv", pid=config["PIDS"]),
         expand("data/processed/{pid}/google_activity_recognition_deltas.csv", pid=config["PIDS"]),
         expand("data/interim/{pid}/phone_valid_sensed_days.csv", pid=config["PIDS"]),
+        expand("data/interim/{pid}/phone_sensed_bins.csv", pid=config["PIDS"]),
         expand("data/processed/{pid}/sms_{sms_type}_{day_segment}.csv",
                             pid=config["PIDS"],
                             sms_type = config["SMS"]["TYPES"],

rules/preprocessing.snakefile

@@ -32,6 +32,16 @@ rule phone_valid_sensed_days:
     script:
         "../src/data/phone_valid_sensed_days.R"
 
+rule phone_sensed_bins:
+    input:
+        all_sensors =  expand("data/raw/{{pid}}/{sensor}_with_datetime.csv", sensor=config["SENSORS"])
+    params:
+        bin_size = config["PHONE_VALID_SENSED_DAYS"]["BIN_SIZE"]
+    output:
+        "data/interim/{pid}/phone_sensed_bins.csv"
+    script:
+        "../src/data/phone_sensed_bins.R"
+
 rule unify_ios_android:
     input:
         sensor_data = "data/raw/{pid}/{sensor}_with_datetime.csv",

src/data/phone_sensed_bins.R

@@ -0,0 +1,31 @@
+source("packrat/init.R")
+
+library(dplyr)
+library(tidyr)
+
+all_sensors <- snakemake@input[["all_sensors"]]
+bin_size <- snakemake@params[["bin_size"]]
+output_file <- snakemake@output[[1]]
+
+# Load all sensors and extract timestamps
+all_sensor_data <- data.frame(timestamp = c())
+for(sensor in all_sensors){
+  sensor_data <- read.csv(sensor, stringsAsFactors = F) %>% 
+    select(local_date, local_hour, local_minute) %>% 
+    mutate(sensor = basename(sensor))
+  all_sensor_data <- rbind(all_sensor_data, sensor_data)
+}
+
+phone_sensed_bins <- all_sensor_data %>% 
+  mutate(bin = (local_minute %/% bin_size) * bin_size) %>% # bin rows into bin_size-minute bins
+  group_by(local_date, local_hour, bin) %>% 
+  summarise(sensor_count = n_distinct(sensor)) %>%
+  ungroup() %>% 
+  complete(nesting(local_date), 
+           local_hour = seq(0, 23, 1), 
+           bin = seq(0, (59 %/% bin_size) * bin_size, bin_size), 
+           fill = list(sensor_count=0)) %>% 
+  pivot_wider(names_from = c(local_hour, bin), values_from = sensor_count)
+
+write.csv(phone_sensed_bins, output_file)
+