upload combined csv file after use

examples/cloudknot_example.py

@@ -119,3 +119,33 @@ def afq_combine_profiles(dummy_argument):
 result_futures2 = knot2.map(["dummy_argument"], job_type="independent")
 result_futures2.result()
 knot2.clobber(clobber_pars=True, clobber_repo=True, clobber_image=True)
+
+# When all jobs are finished, remember to clobber the knot
+# either using the aws console or this function in jupyter notebook:
+result_futures.result()  # waits for futures to resolve, not needed in notebook
+knot.clobber(clobber_pars=True, clobber_repo=True, clobber_image=True)
+
+# we create another knot which downloads the resulting profiles of each subject,
+# combines them into one csv file, and then uploads that csv file
+
+
+def afq_combine_profiles(dummy_argument):
+    from AFQ.api import download_and_combine_afq_profiles
+    import s3fs
+
+    download_and_combine_afq_profiles(
+        "temp", "my_study_bucket", "my_study_prefix/derivatives/afq",
+        upload=s3fs.S3FileSystem())
+
+
+knot2 = ck.Knot(
+    name='afq_combine_subjects-201009-0',
+    func=afq_combine_profiles,
+    base_image='python:3.8',
+    image_github_installs="https://github.com/yeatmanlab/pyAFQ.git",
+    pars_policies=('AmazonS3FullAccess',),
+    bid_percentage=100)
+
+result_futures2 = knot2.map(["dummy_argument"], job_type="independent")
+result_futures2.result()
+knot2.clobber(clobber_pars=True, clobber_repo=True, clobber_image=True)