update app

app.py

@@ -21,8 +21,9 @@ with st.sidebar.expander("📍 Explanation", expanded=False):
 
 with st.sidebar.expander("⚙️ Dataset Options", expanded=False):
     st.markdown("This allows you to switch between the MRPC train and test sets, as well as choose to display only the original paraphrase pairs (MRPC) and/or the corrected pairs (MRPC-R1).")
-    split = st.selectbox("Dataset Split", ["train", "test"])
-    display = st.selectbox("Display only pairs from", ["All", "Only MRPC", "Only MRPC-R1"])
+    split = st.radio("Dataset Split", ["train", "test"])
+    display = st.radio("Display only pairs from", [
+                           "All", "Only MRPC", "Only MRPC-R1"])
 
 ptype = st.sidebar.radio("Display Types", ["All",
                                            "Only Paraphrases (MRPC-R1)",
@@ -31,12 +32,10 @@ ptype = st.sidebar.radio("Display Types", ["All",
                                            "Corrected Paraphrases from MRPC"])
 
 st.sidebar.markdown("**WPD/LD Score Filter Options**")
-filter_by = st.sidebar.selectbox("Filter By Scores From", ["MRPC", "MRPC-R1"])
 display_range_wpd = st.sidebar.slider(
     "Filter by WPD Scores", min_value=0.0, max_value=1.0, value=(0.1, 0.7))
 display_range_ld = st.sidebar.slider(
     "Filter by LD Scores", min_value=0.0, max_value=1.0, value=(0.1, 0.4))
-display_scores = st.sidebar.checkbox("Display scores", value=False)
 
 with st.sidebar.expander("📍 WPD/LD Score Explanation", expanded=False):
     st.markdown("""
@@ -45,9 +44,19 @@ with st.sidebar.expander("📍 WPD/LD Score Explanation", expanded=False):
     * WPD measures difference in the sentence structure
     * LD measures differences in the words used
 
-    By setting WPD to a high range (eg >0.4) and LD to a low range (eg <0.1), we can find paraphrases that do not change much in words used but have very different structures.
+    By setting WPD to a high range (>0.4) and LD to a low range (<0.1), we can find paraphrases that do not change much in words used but have very different structures.
+
+    When LD is set to a high range (>0.5), we can find many pairs labelled as paraphrases in MRPC are not in fact paraphrases.
     """)
 
+    st.markdown("**Additional Options**")
+
+    filter_by = st.radio(
+        "Filter By Scores From", ["MRPC", "MRPC-R1"])
+
+    display_scores = st.checkbox("Display scores", value=False)
+
+
 def load_df(split):
     if split == "train":
         df = pd.read_csv("./mrpc_train_scores.csv")
@@ -114,11 +123,14 @@ def filter_df(df, display, ptype, filter_by, display_scores):
     df_sel.drop(["remarks", "og_label", "new_label"], axis=1, inplace=True)
     return df_sel
 
+
 df = load_df(split)
 
 df_sel = filter_df(df, display, ptype, filter_by, display_scores)
-df_sel.rename(columns={"og_s1": "Original S1 (MRPC)", "og_s2": "Original S2 (MRPC)", "new_s1": "New S1 (MRPC-R1)", "new_s2": "New S2 (MRPC-R1)"}, inplace=True)
+df_sel.rename(columns={"og_s1": "Original S1 (MRPC)", "og_s2": "Original S2 (MRPC)",
+              "new_s1": "New S1 (MRPC-R1)", "new_s2": "New S2 (MRPC-R1)"}, inplace=True)
 
-st.markdown("**MRPC Paraphrase Data Explorer** (Displaying "+str(len(df_sel))+" items)")
+st.markdown("**MRPC Paraphrase Data Explorer** (Displaying " +
+            str(len(df_sel))+" items)")
 
 st.table(data=df_sel)