Merge pull request #11 from scipy-conference/feat/make-runnable

feat: Make notebooks runnable on previous data

Author: guenp

.gitignore

@@ -10,3 +10,5 @@ venv/
 *.ipynb
 # pixi environments
 .pixi
+
+data/

README.md

@@ -13,6 +13,37 @@ The checks and tests are all run using Github actions on every pull request and
 
 This repository is setup for Python 3.11. To customize that, change the `VARIANT` argument in `.devcontainer/devcontainer.json`, change the config options in `.precommit-config.yaml` and change the version number in `.github/workflows/python.yaml`.
 
+## Assign Reviewers
+
+First download the following files from Pretalx into the `data/` directory:
+
+* `scipy_reviewers.csv`  # people who signed up as reviewers
+* `sessions.csv`  # all proposal exported from pretalx
+* `speakers.csv`  # all speakers exported from pretalx
+* `pretalx_reviewers.csv`  # all reviewers copy-pasted from pretalx
+* `scipy_coi_export.csv`  # all responses to the coi form
+* `coi_authors.csv`  # copy pasted values of author names from coi form
+* `tracks.csv`  # manually entered track IDs
+
+Then run the notebooks as Python files in the following order with `pixi`
+
+```
+$ pixi run pre-processing
+$ pixi run assignments
+```
+
+or run the notebooks manually as Jupyter notebooks either by asking for a JupyterLab instance
+
+```
+$ pixi run jupyter lab
+```
+
+or just getting a shell
+
+```
+$ pixi shell
+```
+
 ## Development instructions
 
 ## With devcontainer

assign_reviews.py

@@ -4,6 +4,7 @@
 ####################
 # Imports
 import json
+from pathlib import Path
 
 import numpy as np
 from scipy.optimize import Bounds, LinearConstraint, milp
@@ -120,7 +121,7 @@ def solve_milp(
 ############################
 ## FORMAT AND OUTPUT DATA ##
 ############################
-def format_and_output_result(df_reviewers, df_submissions, solution, post_fix=""):
+def format_and_output_result(df_reviewers, df_submissions, solution, post_fix="", output_dir=Path.cwd() / "output"):
     reviewers = df_reviewers.to_dict("records")
     submissions = df_submissions.to_dict("records")
 
@@ -140,20 +141,20 @@ def format_and_output_result(df_reviewers, df_submissions, solution, post_fix=""
     if DEBUG:
         result = {reviewer["reviewer_id"]: sorted(reviewer["is_tutorial"]) for reviewer in reviewers}
 
-        with open(f"output/review-assignments-debug{post_fix}.json", "w") as fp:
+        with open(output_dir / f"review-assignments-debug{post_fix}.json", "w") as fp:
             fp.write(json.dumps(result, indent=4))
 
     result = {reviewer["reviewer_id"]: reviewer["assigned_submission_ids"] for reviewer in reviewers}
 
-    with open(f"output/review-assignments{post_fix}.json", "w") as fp:
+    with open(output_dir / f"review-assignments{post_fix}.json", "w") as fp:
         fp.write(json.dumps(result, indent=4))
 
     for submission, assignments in zip(submissions, solution.T):
         submission["assigned_reviewer_ids"] = df_reviewers.reviewer_id[assignments].values.tolist()
 
     result = {submission["submission_id"]: submission["assigned_reviewer_ids"] for submission in submissions}
 
-    with open(f"output/submission-assignments{post_fix}.json", "w") as fp:
+    with open(output_dir / f"submission-assignments{post_fix}.json", "w") as fp:
         fp.write(json.dumps(result, indent=4))
 
     return reviewers, submissions

data/.gitkeep

@@ -19,29 +19,34 @@
 #     name: python
 #     nbconvert_exporter: python
 #     pygments_lexer: ipython3
-#     version: 3.12.1
+#     version: 3.12.2
 # ---
 
 # %%
+from pathlib import Path
+
 import duckdb
 from IPython import display
 
+# %%
+data_dir = Path.cwd() / ".." / "data"
+
 # Raw data to import
 raw_files = dict(
-    scipy_reviewers="../data/scipy_reviewers.csv",  # people who signed up as reviewers
-    pretalx_sessions="../data/sessions.csv",  # all proposal exported from pretalx
-    pretalx_speakers="../data/speakers.csv",  # all speakers exported from pretalx
-    pretalx_reviewers="../data/pretalx_reviewers.csv",  # all reviewers copy-pasted from pretalx
-    coi_reviewers="../data/scipy_coi_export.csv",  # all responses to the coi form
-    coi_authors="../data/coi_authors.csv",  # copy pasted values of author names from coi form
-    tracks="../data/tracks.csv",  # manually entered track IDs
+    scipy_reviewers=data_dir / "scipy_reviewers.csv",  # people who signed up as reviewers
+    pretalx_sessions=data_dir / "sessions.csv",  # all proposal exported from pretalx
+    pretalx_speakers=data_dir / "speakers.csv",  # all speakers exported from pretalx
+    pretalx_reviewers=data_dir / "pretalx_reviewers.csv",  # all reviewers copy-pasted from pretalx
+    coi_reviewers=data_dir / "scipy_coi_export.csv",  # all responses to the coi form
+    coi_authors=data_dir / "coi_authors.csv",  # copy pasted values of author names from coi form
+    tracks=data_dir / "tracks.csv",  # manually entered track IDs
 )
 
 # Output
-database_file = "../data/assign_reviews.db"
+database_file = data_dir / "assign_reviews.db"
 
 # %%
-con = duckdb.connect(database_file)
+con = duckdb.connect(str(database_file))
 
 
 # %%
@@ -141,7 +146,7 @@ def create_and_show_table(file_name, table_name, show=True):
 # Reviewers who signed up for pretalx but did not fill in COI
 
 # %%
-con = duckdb.connect(database_file)
+con = duckdb.connect(str(database_file))
 
 # %%
 df = con.sql(
@@ -359,7 +364,7 @@ def create_and_show_table(file_name, table_name, show=True):
 con.sql("table reviewers_with_tracks").df()
 
 # %%
-con.sql("select email as reviewer_id, list(track_id) as tracks from reviewers_with_tracks group by email")
+con.sql("select email as reviewer_id, list(track_ids) as tracks from reviewers_with_tracks group by email")
 
 # %% [markdown]
 # # Final tables for script
@@ -411,5 +416,3 @@ def create_and_show_table(file_name, table_name, show=True):
 
 # %%
 con.close()
-
-# %%

notebooks/pre-processing.py

@@ -19,16 +19,16 @@
 #     name: python
 #     nbconvert_exporter: python
 #     pygments_lexer: ipython3
-#     version: 3.12.1
+#     version: 3.12.2
 # ---
 
 # %%
 ####################
 ## ASSIGN REVIEWS ##
 ####################
-# Imports
 import json
 import sys
+from pathlib import Path
 
 import duckdb
 import pandas as pd
@@ -40,16 +40,18 @@
 # # Start script
 
 # %%
-# mkdir output
+data_dir = Path().cwd() / ".." / "data"
+output_dir = Path().cwd() / ".." / "output"
+output_dir.mkdir(exist_ok=True)
 
 # %%
 ASSIGN_TUTORIALS_TO_ANYONE = False
 TUTORIAL_COEFF = 0.8
 
 DEBUG = True
 
-database_file = "../data/assign_reviews.db"
-con = duckdb.connect(database_file)
+database_file = data_dir / "assign_reviews.db"
+con = duckdb.connect(str(database_file))
 df_submissions = con.sql("table submissions_to_assign").df()
 df_reviewers = con.sql("table reviewers_to_assign").df()
 
@@ -82,7 +84,9 @@
     TUTORIAL_COEFF,
     ASSIGN_TUTORIALS_TO_ANYONE,
 )
-reviewers, submissions = format_and_output_result(df_reviewers, df_submissions_tutorials, solution, post_fix="00")
+reviewers, submissions = format_and_output_result(
+    df_reviewers, df_submissions_tutorials, solution, post_fix="00", output_dir=output_dir
+)
 
 # %%
 df = pd.DataFrame(reviewers)
@@ -141,7 +145,7 @@
 )
 if solution is not None:
     reviewers, submissions = format_and_output_result(
-        df_reviewers_no_submissions, df_submissions_no_tutorials, solution, post_fix="01"
+        df_reviewers_no_submissions, df_submissions_no_tutorials, solution, post_fix="01", output_dir=output_dir
     )
 
 # %%
@@ -217,7 +221,7 @@
 
 if solution is not None:
     reviewers, submissions = format_and_output_result(
-        df_reviewers_only_tut, df_submissions_few_reviewers, solution, post_fix="02"
+        df_reviewers_only_tut, df_submissions_few_reviewers, solution, post_fix="02", output_dir=output_dir
     )
 
 # %%
@@ -314,17 +318,18 @@
 # ## Final export
 
 # %%
-database_file = "../data/assign_reviews.db"
-con = duckdb.connect(database_file)
+database_file = data_dir / "assign_reviews.db"
+con = duckdb.connect(str(database_file))
 
 # %%
 reviewer_assignments_final = {
-    item["reviewer_id"]: item["assigned_submission_ids"]
+    item["reviewer_id"]: item["assigned_submission_ids"].tolist()
     for item in con.sql("table reviewer_assignments_02")
     .df()[["reviewer_id", "assigned_submission_ids"]]
     .to_dict("records")
 }
-with open("output/reviewer-assignments.json", "w") as fp:
+
+with open(output_dir / "reviewer-assignments.json", "w") as fp:
     fp.write(json.dumps(reviewer_assignments_final, indent=4))
 
 # %%

notebooks/run-assignments.py

@@ -10,6 +10,8 @@ channels = ["conda-forge"]
 platforms = ["linux-64", "osx-64", "osx-arm64"]
 
 [tasks]
+pre-processing = "cd notebooks && python pre-processing.py"
+assignments = "cd notebooks && python run-assignments.py"
 
 [dependencies]
 python = "3.12.*"