From 44bbe1743558f6e42021c4276283ee734cde7c22 Mon Sep 17 00:00:00 2001
From: JulioV <JulioV@users.noreply.github.com>
Date: Wed, 13 Nov 2019 14:54:24 -0500
Subject: [PATCH] Update location_barnett_metrics.R to handle empty dataframes
 and location sequences without flights

---
 src/features/location_barnett_metrics.R | 40 ++++++++++++++++++++-----
 1 file changed, 33 insertions(+), 7 deletions(-)

diff --git a/src/features/location_barnett_metrics.R b/src/features/location_barnett_metrics.R
index d8aae256..89e18bf3 100644
--- a/src/features/location_barnett_metrics.R
+++ b/src/features/location_barnett_metrics.R
@@ -2,6 +2,26 @@ source("packrat/init.R")
 
 library(dplyr)
 
+write_empty_file <- function(file_path){
+  write.csv(data.frame(local_date= character(), 
+                        hometime= numeric(), 
+                        disttravelled= numeric(), 
+                        rog= numeric(), 
+                        maxdiam= numeric(), 
+                        maxhomedist= numeric(), 
+                        siglocsvisited= numeric(), 
+                        avgflightlen= numeric(), 
+                        stdflightlen= numeric(), 
+                        avgflightdur= numeric(), 
+                        stdflightdur= numeric(), 
+                        probpause= numeric(), 
+                        siglocentropy= numeric(), 
+                        minsmissing= numeric(), 
+                        circdnrtn= numeric(), 
+                        wkenddayrtn= numeric()
+                      ), file_path, row.names = F)
+}
+
 # Load Ian Barnett's code. Taken from https://scholar.harvard.edu/ibarnett/software/gpsmobility
 file.sources = list.files(c("src/features/location_barnett"), pattern="*.R$", full.names=TRUE, ignore.case=TRUE)
 sapply(file.sources,source,.GlobalEnv)
@@ -12,13 +32,19 @@ timezone <- snakemake@params[["timezone"]]
 location <- read.csv(snakemake@input[[1]], stringsAsFactors = F) %>%
   select(timestamp, latitude = double_latitude, longitude = double_longitude, altitude = double_altitude, accuracy)
 
-if (nrow(location) > 0){
+if (nrow(location) > 1){
     features <- MobilityFeatures(location, ACCURACY_LIM = accuracy_limit, tz = timezone)
-
-    # Copy index (dates) as a column 
-    outmatrix = cbind(rownames(features$featavg), features$featavg)
-    colnames(outmatrix)=c("local_date",tolower(colnames(features$featavg)))
-    write.csv(outmatrix,snakemake@output[[1]], row.names = F)
+    if(is.null(features)){
+      write_empty_file(snakemake@output[[1]])
+    } else{
+      # Copy index (dates) as a column 
+      outmatrix <- cbind(rownames(features$featavg), features$featavg)
+      outmatrix <- as.data.frame(outmatrix)
+      outmatrix[-1] <- lapply(lapply(outmatrix[-1], as.character), as.numeric)
+      colnames(outmatrix)=c("local_date",tolower(colnames(features$featavg)))
+      write.csv(outmatrix,snakemake@output[[1]], row.names = F)
+    }
+    
 } else {
-    write.csv(data.frame(),snakemake@output[[1]])
+    write_empty_file(snakemake@output[[1]])    
 }
\ No newline at end of file