assets

pages/2_Statistics.py

@@ -1,12 +1,10 @@
 import pandas as pd
 import streamlit as st
 
-best_results_df = pd.read_csv("output/best_tasks_with_hits.csv")
+best_results_df = pd.read_csv("assets/best_tasks_with_hits.csv")
 
 
-worst_results_df = pd.read_csv(
-    "output/worst_tasks_with_hits.csv"
-)  # , data_path="output/papers_with_dependencies.csv",
+worst_results_df = pd.read_csv("assets/worst_tasks_with_hits.csv")
 
 show_worst_best_statistics = st.sidebar.checkbox(
     label="show worst/best statistics grouped by area"
@@ -16,11 +14,13 @@ show_area_aggregated_results = st.sidebar.checkbox(
     label="show results aggregated by area"
 )
 if show_worst_best_statistics:
-    st.markdown("""
+    st.markdown(
+        """
     ## Worst/best queries
     The following are top 10 worst/best queries per area by number of hits.
     There are at least 10 documents per query in the test set, so number of hits/10 is the accuracy.
-    """)
+    """
+    )
     sort_key = st.selectbox("sort by", list(best_results_df.columns))
     st.markdown("## Queries with best results")
     st.table(best_results_df.sort_values(sort_key, ascending=False))
@@ -36,4 +36,3 @@ if show_area_aggregated_results:
     st.table(best_results_agg.sort_values(sort_key, ascending=False))
     st.markdown("Worst results")
     st.table(worst_results_agg.sort_values(sort_key, ascending=False))
-