From a4ad1a16abd7a8bf204d76b89e172d12ae1e1389 Mon Sep 17 00:00:00 2001
From: Arya Massarat <23412689+aryarm@users.noreply.github.com>
Date: Sat, 11 Sep 2021 08:55:14 -0700
Subject: [PATCH] remove threads directive from each rule. every thread should
 get a single core

---
 Snakefile  | 10 ----------
 config.yml |  8 --------
 2 files changed, 18 deletions(-)

diff --git a/Snakefile b/Snakefile
index b3eab9f..0254a48 100755
--- a/Snakefile
+++ b/Snakefile
@@ -112,7 +112,6 @@ rule stitch:
         files = directory(config['out']+"/{sample}/stitch"+("-lowQual" if check_config('low_qual_ortho', check_config('parallel')) else "")+"/stitched.files")
     conda: "envs/default.yml"
     benchmark: config['out']+"/{sample}/benchmark/stitch"+("-lowQual" if check_config('low_qual_ortho', check_config('parallel')) else "")+".tsv"
-    threads: config['cores']
     shell:
         "scripts/stitch.py {params} {input} {output.project}"
 
@@ -124,7 +123,6 @@ rule export_ortho:
         str(Path(rules.stitch.output.project).parents[0])+"/ortho.tiff"
     conda: "envs/default.yml"
     benchmark: config['out']+"/{sample}/benchmark/export_ortho.tsv"
-    threads: config['cores']
     shell:
         "scripts/export_ortho.py {input} {output}"
 
@@ -139,7 +137,6 @@ rule segment:
         low = config['out']+"/{sample}/segments/low/{image}.json" if check_config('parallel') else config['out']+"/{sample}/segments/low.json"
     conda: "envs/default.yml"
     benchmark: config['out']+"/{sample}/benchmark/segments/"+("{image}" if check_config('parallel') else "ortho")+".tsv"
-    threads: config['cores']
     shell:
         "scripts/segment.py {params} {input} {output}"
 
@@ -154,7 +151,6 @@ rule transform:
         confidence="(high|low)"
     conda: "envs/default.yml"
     benchmark: config['out']+"/{sample}/benchmark/transform/{confidence}-{image}.json"
-    threads: config['cores']
     shell:
         "scripts/transform.py {input} {output}"
 
@@ -192,7 +188,6 @@ rule watershed:
         segments = config['out']+"/{sample}/segments"+exp_str()+".json"
     conda: "envs/default.yml"
     benchmark: config['out']+"/{sample}/benchmark/watershed"+exp_str()+".tsv"
-    threads: config['cores']
     shell:
         "scripts/watershed.py {input.ortho} {params.high_dir} {params.low_dir} {output.segments}"
 
@@ -283,7 +278,6 @@ rule train:
         config['out']+"/{sample}/train"+exp_str()+"/model.rda",
         config['out']+"/{sample}/train"+exp_str()+"/variable_importance.tsv"
     conda: "envs/classify.yml"
-    threads: config['cores']
     shell:
         "Rscript scripts/classify_train.R {input} {output}"
 
@@ -327,7 +321,6 @@ rule classify:
         config['out']+"/{sample}/classify"+exp_str()+"/{image}.tsv"
     conda: "envs/classify.yml"
     benchmark: config['out']+"/{sample}/benchmark/classify"+exp_str()+"/{image}.tsv"
-    threads: config['cores']
     shell:
         "Rscript scripts/classify_test.R {input} {output}"
 
@@ -338,7 +331,6 @@ rule test:
         config['out']+"/{sample}/test"+exp_str()+"/classify/{image}.tsv"
     conda: "envs/classify.yml"
     benchmark: config['out']+"/{sample}/benchmark/test"+exp_str()+"/classify/{image}.tsv"
-    threads: config['cores']
     shell:
         "Rscript scripts/classify_test.R {input} {output}"
 
@@ -434,7 +426,6 @@ rule segments_map:
         config['out']+"/{sample}/segments-map"+exp_str()+".tiff"
     conda: "envs/default.yml"
     benchmark: config['out']+"/{sample}/benchmark/segments-map"+exp_str()+".tsv"
-    threads: config['cores']
     shell:
         "scripts/map.py -l {input.img} {input.labels} {output}"
 
@@ -448,7 +439,6 @@ rule map:
         config['out']+"/{sample}/map"+exp_str()+".tiff"
     conda: "envs/default.yml"
     benchmark: config['out']+"/{sample}/benchmark/map"+exp_str()+".tsv"
-    threads: config['cores']
     shell:
         "scripts/map.py {input.img} {input.labels} {output} {input.predicts}"
 
diff --git a/config.yml b/config.yml
index aab4a89..4adba97 100644
--- a/config.yml
+++ b/config.yml
@@ -70,14 +70,6 @@ model: data/models/test-all-exp.rda
 # Defaults to 'out' if not provided
 out: out
 
-# specifying the number of cores used to process each dataset
-# TODO: note that this wouldn't work yet - we need to have parallel environments defined
-# in purves first before we can use snakemake's multithread feature
-# I just have it defined here so in the future if we have parallel environment
-# we could use it
-cores: 5        # The number of cores to use for this step
-
-
 # FOR THE RULE extract_images (only uncomment if using that rule)
 # specifying the list of segment labels that you want to extract the
 # source images for