add the rest of the code.

[articlequality_ordinal.git] / Makefile

diff --git a/Makefile b/Makefile
new file mode 100644 (file)
index 0000000..a9cb94a
--- /dev/null
+++ b/Makefile
@@ -0,0 +1,37 @@
+SHELL:=/bin/bash
+
+data/20200301_article_labelings.json_SUCCESS:
+       ./run_aql_jobs.sh
+
+data/20200301_article_labelings_sample.json:sample_training_labels.py
+       source ./bin/activate; \
+       ./sample_training_labels.py
+
+
+data/article_sample.csv:sample_articles.py
+       source ./bin/activate; \
+       start_spark_and_run.sh 1 sample_articles.py
+
+data/scored_article_sample.feather:data/article_sample_set.csv ores_scores_sample.py
+       source ./bin/activate; \
+       python3 ores_scores_sample.py data/article_sample_set.parquet data/scored_article_sample.feather
+
+# run this step on kibo
+data/20200301_al_sample_revisions.w_text.json:data/20200301_article_labelings_sample.json
+       source ./bin/activate; \
+       python3 articlequality/utility fetch_text \
+         --api-host=https://en.wikipedia.org \
+         --labelings=data/20200301_article_labelings_sample.json \
+         --output=data/20200301_al_sample_revisions.w_text.json \
+
+# run this step on kibo
+data/20200301_al_sample_revisions.w_scores.json:data/20200301_al_sample_revisions.w_text.json
+       python3 score_sample_labels.py
+
+models/ordinal_quality.RDS:data/20200301_al_sample_revisions.w_text.json ordinal_quality_models.R
+       Rscript ordinal_quality_models.R
+
+
+
+
+PHONY: data/20200301_article_labelings.json