ICML2023_papers / paper_list.py
hysts's picture
hysts HF staff
Update
8235d81
from __future__ import annotations
import numpy as np
import pandas as pd
class PaperList:
def __init__(self):
self.organization_name = "ICML2023"
self.table = pd.read_csv("papers.csv")
self._preprocess_table()
self.table_header = """
<tr>
<td width="38%">Title</td>
<td width="25%">Authors</td>
<td width="5%">arXiv</td>
<td width="5%">GitHub</td>
<td width="7%">Paper pages</td>
<td width="5%">Spaces</td>
<td width="5%">Models</td>
<td width="5%">Datasets</td>
<td width="5%">Claimed</td>
</tr>"""
def _preprocess_table(self) -> None:
self.table["title_lowercase"] = self.table.title.str.lower()
rows = []
for row in self.table.itertuples():
title = f"{row.title}"
arxiv = f'<a href="{row.arxiv}" target="_blank">arXiv</a>' if isinstance(row.arxiv, str) else ""
github = f'<a href="{row.github}" target="_blank">GitHub</a>' if isinstance(row.github, str) else ""
hf_paper = (
f'<a href="{row.hf_paper}" target="_blank">Paper page</a>' if isinstance(row.hf_paper, str) else ""
)
hf_space = f'<a href="{row.hf_space}" target="_blank">Space</a>' if isinstance(row.hf_space, str) else ""
hf_model = f'<a href="{row.hf_model}" target="_blank">Model</a>' if isinstance(row.hf_model, str) else ""
hf_dataset = (
f'<a href="{row.hf_dataset}" target="_blank">Dataset</a>' if isinstance(row.hf_dataset, str) else ""
)
author_linked = "✅" if ~np.isnan(row.n_linked_authors) and row.n_linked_authors > 0 else ""
n_linked_authors = "" if np.isnan(row.n_linked_authors) else int(row.n_linked_authors)
n_authors = "" if np.isnan(row.n_authors) else int(row.n_authors)
claimed_paper = "" if n_linked_authors == "" else f"{n_linked_authors}/{n_authors} {author_linked}"
row = f"""
<tr>
<td>{title}</td>
<td>{row.authors}</td>
<td>{arxiv}</td>
<td>{github}</td>
<td>{hf_paper}</td>
<td>{hf_space}</td>
<td>{hf_model}</td>
<td>{hf_dataset}</td>
<td>{claimed_paper}</td>
</tr>"""
rows.append(row)
self.table["html_table_content"] = rows
def render(self, search_query: str, case_sensitive: bool, filter_names: list[str]) -> tuple[str, str]:
df = self.table
if search_query:
if case_sensitive:
df = df[df.title.str.contains(search_query)]
else:
df = df[df.title_lowercase.str.contains(search_query.lower())]
has_arxiv = "arXiv" in filter_names
has_github = "GitHub" in filter_names
has_hf_space = "Space" in filter_names
has_hf_model = "Model" in filter_names
has_hf_dataset = "Dataset" in filter_names
df = self.filter_table(df, has_arxiv, has_github, has_hf_space, has_hf_model, has_hf_dataset)
n_claimed = len(df[df.n_linked_authors > 0])
return f"{len(df)} ({n_claimed} claimed)", self.to_html(df, self.table_header)
@staticmethod
def filter_table(
df: pd.DataFrame,
has_arxiv: bool,
has_github: bool,
has_hf_space: bool,
has_hf_model: bool,
has_hf_dataset: bool,
) -> pd.DataFrame:
if has_arxiv:
df = df[~df.arxiv.isna()]
if has_github:
df = df[~df.github.isna()]
if has_hf_space:
df = df[~df.hf_space.isna()]
if has_hf_model:
df = df[~df.hf_model.isna()]
if has_hf_dataset:
df = df[~df.hf_dataset.isna()]
return df
@staticmethod
def to_html(df: pd.DataFrame, table_header: str) -> str:
table_data = "".join(df.html_table_content)
html = f"""
<table>
{table_header}
{table_data}
</table>"""
return html