diff --git a/gpextras.py b/gpextras.py
index f4c8722..7fd72be 100644
--- a/gpextras.py
+++ b/gpextras.py
@@ -105,18 +105,18 @@ def simpleMapCommands(self,task,gp,mapCom,src,dst):
     def simpleMapReduceCommands(self,task,gp,mapCom,reduceCom,src,dst):
         """A map-reduce job with one input."""
         p = task.reduceParallel(gp)
-        return [ "%s --input %s --output %s --mapper '%s'  --numReduceTasks %d --reducer '%s'" \
-                 % (self.mrsCommand,src,dst,mapCom,p,reduceCom) ]
+        return [ "%s --input %s --output %s --numReduceTasks %d --mapper '%s'  --reducer '%s'" \
+                 % (self.mrsCommand,src,dst,p,mapCom,reduceCom) ]
 
     def joinCommands(self,task,gp,mapComs,reduceCom,srcs,midpoint,dst):
         """A map-reduce job with several inputs."""
         p = task.reduceParallel(gp)
-        def midi(i): return midpoint + '-' + str(i)
+        def mid(i): return midpoint + '-' + str(i)
         subplan = []
         for i in range(len(srcs)):
             subplan.append("%s --input %s --output %s --mapper '%s'" \
                            % (self.mrsCommand,srcs[i],mid(i),mapComs[i]))
         allMidpoints = ",".join([mid(i) for i in range(len(srcs))])
-        subplan.append("%s --inputs %s --output %s --mapper cat --numReduceTasks %d --reducer '%s'" \
+        subplan.append("%s --inputs %s --output %s --numReduceTasks %d --mapper cat --reducer '%s'" \
                        % (self.mrsCommand,allMidpoints,dst,p,reduceCom))
         return subplan
diff --git a/mrs_gp.py b/mrs_gp.py
index 923b3cf..c6de859 100644
--- a/mrs_gp.py
+++ b/mrs_gp.py
@@ -686,9 +686,12 @@ def setupFiles(indirList,outdir,numReduceTasks):
     single output, it will also be a file.
     """
     #clear space/make directory for output, if necessary
-    if os.path.exists(outdir):
+    if os.path.exists(outdir):  
         logging.warn('removing %s' % (outdir))
-        shutil.rmtree(outdir)
+        if os.path.isdir(outdir):
+            shutil.rmtree(outdir)
+        else:
+            os.remove(outdir)
 
     indirs = []
     infiles = []
diff --git a/mrs_test/Makefile b/mrs_test/Makefile
index f20f7a5..c9ea1cf 100644
--- a/mrs_test/Makefile
+++ b/mrs_test/Makefile
@@ -4,7 +4,7 @@ ASYNC = 0
 
 #
 clean: 
-	rm -rf tmp/* parallel-events.txt
+	rm -rf tmp tmp.txt sharded parallel-events.txt server.log test test2 gpig_views
 
 backup:
 	cp ../mrs_gp_v4.py ../backup/`date "+%H:%M:%S"`-mrs_gp_v4.py
diff --git a/mrs_test/Makefile.stress b/mrs_test/Makefile.stress
deleted file mode 100644
index 76fb3d6..0000000
--- a/mrs_test/Makefile.stress
+++ /dev/null
@@ -1,79 +0,0 @@
-#comments: with full_test.txt, about 100M
-#  disk and ram times are comparable
-#  ram is fastest w/o parallelism (K=1), then it's about 2x speed of disk
-# with full_train.txt, about 1G
-#  ram is much faster
-#  not much speedup w parallelism (K=1 is best)
-
-
-RAM=/mnt/ramdisk/mrs/
-K=60
-BIGFILE=/afs/cs.cmu.edu/user/wcohen/bigML/RCV1/RCV1.full_test.txt
-#BIGFILE=/afs/cs.cmu.edu/user/wcohen/bigML/RCV1/RCV1.full_train.txt
-DIR=$(RAM)
-NBCOUNT='python streamNaiveBayesLearner.py --streamTrain 100'
-#NBCOUNT='python streamNaiveBayesLearner.py --train'
-SUM='python sum-events.py'
-
-MRS=pypy ../mrs_gp.py
-
-simple-disk-times:
-	echo == disk stream - about 3 sec for 100M, 2:03 for 1G
-	time cat $(BIGFILE) | python fixrcv.py > bigfile1/test.txt
-	echo == sharding disk-disk - about 10-11 sec for 100M with K=10, 1:18 for 1G with K=5
-	time $(MRS) --input bigfile1 --output bigfile1-sharded --mapper cat --reducer cat --numReduceTasks $(K)
-	echo == copy sharded disk-disk - about 8-10 sec	for 100M with K=10, 1:16 for 1G with K=5
-	time $(MRS) --input bigfile1-sharded --output bigfile1-copy --mapper cat --reducer cat --numReduceTasks $(K)
-
-simple-ram-times:
-	echo == disk to ram stream - about 3 sec for 100M, 3:19 for 1G with K=5
-	time cat /afs/cs.cmu.edu/user/wcohen/bigML/RCV1/RCV1.full_test.txt | python fixrcv.py > $(RAM)/bigfile1/test.txt
-	echo == sharding ram-ram - about 10-11 sec for 100M , 0:08 for 1G with K=5
-	time $(MRS) --input $(RAM)/bigfile1 --output $(RAM)/bigfile1-sharded --mapper cat --reducer cat --numReduceTasks $(K)
-	echo == copy sharded ram-ram - about 8-10 sec for 100M, 0:07 for 1G with K=5	
-	time $(MRS) --input $(RAM)/bigfile1-sharded --output $(RAM)/bigfile1-copy --mapper cat --reducer cat --numReduceTasks $(K)        
-	echo == resharded ram-ram - about 8.7 sec for 100M with k=30
-	time $(MRS) --input $(RAM)/bigfile1-sharded --output $(RAM)/bigfile1-copy --mapper cat --reducer cat --numReduceTasks `expr $(K) / 2`
-
-# for train data - 1G 
-# k	shard	copy
-# 1	5.7	5.4   
-# 3	7.7	7.3	
-# 5	7.9	7.5
-# 10	9.0	8.2
-# 20	9.7	8.8	
-
-simple-ram-times-sweep-k:
-	time $(MRS) --input $(RAM)/bigfile1 --output $(RAM)/bigfile1-sharded --mapper cat --reducer cat --numReduceTasks 1
-	time $(MRS) --input $(RAM)/bigfile1-sharded --output $(RAM)/bigfile1-copy --mapper cat --reducer cat --numReduceTasks 1
-	time $(MRS) --input $(RAM)/bigfile1 --output $(RAM)/bigfile1-sharded --mapper cat --reducer cat --numReduceTasks 3
-	time $(MRS) --input $(RAM)/bigfile1-sharded --output $(RAM)/bigfile1-copy --mapper cat --reducer cat --numReduceTasks 3
-	time $(MRS) --input $(RAM)/bigfile1 --output $(RAM)/bigfile1-sharded --mapper cat --reducer cat --numReduceTasks 5
-	time $(MRS) --input $(RAM)/bigfile1-sharded --output $(RAM)/bigfile1-copy --mapper cat --reducer cat --numReduceTasks 5
-	time $(MRS) --input $(RAM)/bigfile1 --output $(RAM)/bigfile1-sharded --mapper cat --reducer cat --numReduceTasks 10
-	time $(MRS) --input $(RAM)/bigfile1-sharded --output $(RAM)/bigfile1-copy --mapper cat --reducer cat --numReduceTasks 10
-	time $(MRS) --input $(RAM)/bigfile1 --output $(RAM)/bigfile1-sharded --mapper cat --reducer cat --numReduceTasks 20
-	time $(MRS) --input $(RAM)/bigfile1-sharded --output $(RAM)/bigfile1-copy --mapper cat --reducer cat --numReduceTasks 20
-
-# for test data - 100M --train
-#      version 1 	  version 2  	    pypy v2
-# k	shard	learn	shard	  learn	    shard	learn
-# 1	5.7	 3:44	2.8	  3:05	    1:27	3:46
-# 3	6.7	 8:12	3.0	  3:38	    1:28	7:34
-# 10	8.5	14:02	
-# 20	9.0	20:40	3.5	  7:10	    1:30	26:32
-
-# for test data - 100M --streamTrain 100 pypy
-# k	shard	learn	%learnCPU
-# 1	1:31	  26:49		102
-# 3	1:29	1:01:35		250
-# 10
-# 20
-
-learn-ram-times:
-	echo sharding
-	$(MRS) --serve
-	time $(MRS) --input $(RAM)/bigfile1 --output $(RAM)/bigfile1-sharded --mapper cat --reducer cat --numReduceTasks $(K)
-	echo learning
-	time $(MRS) --input $(RAM)/bigfile1-sharded --output $(RAM)/events --numReduceTasks $(K) --mapper $(NBCOUNT) --reducer $(SUM)
-	$(MRS) --shutdown
diff --git a/mrs_test/test/small.txt b/mrs_test/test/small.txt
deleted file mode 100644
index b29f9a6..0000000
--- a/mrs_test/test/small.txt
+++ /dev/null
@@ -1,20 +0,0 @@
-input 1
-input 2
-input 3
-input 4
-input 5
-input 6
-input 7
-input 8
-input 9
-input 10
-input 11
-input 12
-input 13
-input 14
-input 15
-input 16
-input 17
-input 18
-input 19
-input 20
diff --git a/mrs_test/test2/small.txt b/mrs_test/test2/small.txt
deleted file mode 100644
index 7ed0fb2..0000000
--- a/mrs_test/test2/small.txt
+++ /dev/null
@@ -1,20 +0,0 @@
-secondary-input 1
-secondary-input 2
-secondary-input 3
-secondary-input 4
-secondary-input 5
-secondary-input 6
-secondary-input 7
-secondary-input 8
-secondary-input 9
-secondary-input 10
-secondary-input 11
-secondary-input 12
-secondary-input 13
-secondary-input 14
-secondary-input 15
-secondary-input 16
-secondary-input 17
-secondary-input 18
-secondary-input 19
-secondary-input 20