Upload gwf_app.py

gwf_app.py

@@ -0,0 +1,93 @@
+import os
+import re
+import numpy as np
+import pandas as pd
+import plotly.express as px
+import streamlit as st
+
+st.set_page_config(layout="wide")
+DATA_FILE = "data/gwf_2017-2021_specter2_base.json"
+THEMES = {"cluster": "fall", "year": "mint", "source": "phase"}
+
+
+def load_df(data_file: os.PathLike):
+    df = pd.read_json(data_file, orient="records")
+    df["x"] = df["point2d"].apply(lambda x: x[0])
+    df["y"] = df["point2d"].apply(lambda x: x[1])
+    df["year"] = df["year"].replace("", 0)
+    df["year"] = df["year"].astype(int)
+    if "publication_type" in df.columns:
+        df["type"] = df["publication_type"]
+        df = df.drop(columns=["point2d", "publication_type"])
+    else:
+        df = df.drop(columns=["point2d"])
+    return df
+
+
+@st.cache_data
+def load_dataframe():
+    return load_df(DATA_FILE)
+
+
+DF = load_dataframe()
+DF["opacity"] = 0.04
+min_year, max_year = DF[DF["year"] > 0]["year"].min(), DF[DF["year"] > 0]["year"].max()
+
+with st.sidebar:
+    start_year, end_year = st.select_slider(
+        "Publication year",
+        options=[str(y) for y in range(min_year, max_year + 1)],
+        value=(str(min_year), str(max_year)),
+    )
+    src = st.text_input("Source")
+
+    author_names = st.text_input("Author names (separated by comma)")
+
+    title = st.text_input("Title")
+
+    start_year = int(start_year)
+    end_year = int(end_year)
+    df_mask = (DF["year"] >= start_year) & (DF["year"] <= end_year)
+
+    if src:
+        df_mask = df_mask & DF.source.apply(lambda x: src.lower() in x.lower())
+
+    if author_names:
+        authors = [a.strip() for a in author_names.split(",")]
+        author_mask = DF.authors.apply(
+            lambda row: all(any(re.match(rf".*{a}.*", x, re.IGNORECASE) for x in row) for a in authors)
+        )
+        df_mask = df_mask & author_mask
+
+    if title:
+        df_mask = df_mask & DF.title.apply(lambda x: title.lower() in x.lower())
+
+    DF.loc[df_mask, "opacity"] = 1.0
+    st.write(f"Number of points: {DF[df_mask].shape[0]}")
+
+    color = st.selectbox("Color", ("cluster", "year", "source"))
+
+
+fig = px.scatter(
+    DF,
+    x="x",
+    y="y",
+    opacity=DF["opacity"],
+    color=color,
+    width=1000,
+    height=800,
+    hover_data=["title", "authors", "year", "source", "type"],
+    color_continuous_scale=THEMES[color],
+)
+fig.update_layout(
+    # margin=dict(l=10, r=10, t=10, b=10),
+    showlegend=False,
+    font=dict(
+        family="Times New Roman",
+        size=30,
+    ),
+)
+fig.update_xaxes(title="")
+fig.update_yaxes(title="")
+
+st.plotly_chart(fig, use_container_width=True)