add accept and reject button and interaction

bdikit/visualization/scope_reducing.py

@@ -1,4 +1,5 @@
 import json
+import logging
 
 import altair as alt
 import numpy as np
@@ -11,6 +12,8 @@
 pn.extension("mathjax")
 pn.extension("vega")
 
+logger = logging.getLogger(__name__)
+
 
 class SRHeatMapManager:
     def __init__(self) -> None:
@@ -23,6 +26,9 @@ def __init__(self) -> None:
         self.rec_cols_gdc = None
         self.clusters = None
 
+        # Selected column
+        self.selected_row = None
+
     def _load_json(self):
         with open(self.json_path) as f:
             data = json.load(f)
@@ -32,7 +38,8 @@ def _write_json(self, data):
         with open(self.json_path, "w") as f:
             json.dump(data, f)
 
-    def get_heatmap(self, recommendations):
+    def get_heatmap(self):
+        recommendations = self._load_json()
         rec_cols = set()
         rec_table = []
         rec_list = []
@@ -124,6 +131,48 @@ def _get_column_values(self, properties):
         else:
             return None
 
+    def _accept_match(self, col_name=None, match_name=None):
+        if self.selected_row is None:
+            return
+        col_name = self.selected_row["Column"].values[0]
+        match_name = self.selected_row["Recommendation"].values[0]
+        recommendations = self._load_json()
+        for idx, d in enumerate(recommendations):
+            candidate_name = d["Candidate column"]
+            if candidate_name != col_name:
+                continue
+            for top_k_name, top_k_score in d["Top k columns"]:
+                if top_k_name == match_name:
+                    recommendations[idx] = {
+                        "Candidate column": candidate_name,
+                        "Top k columns": [[top_k_name, top_k_score]],
+                    }
+                    self._write_json(recommendations)
+                    self.get_heatmap()
+                    return
+
+    def _reject_match(self):
+        if self.selected_row is None:
+            return
+        col_name = self.selected_row["Column"].values[0]
+        match_name = self.selected_row["Recommendation"].values[0]
+        recommendations = self._load_json()
+        for idx, d in enumerate(recommendations):
+            candidate_name = d["Candidate column"]
+            if candidate_name != col_name:
+                continue
+            new_top_k = []
+            for top_k_name, top_k_score in d["Top k columns"]:
+                if top_k_name != match_name:
+                    new_top_k.append([top_k_name, top_k_score])
+            recommendations[idx] = {
+                "Candidate column": candidate_name,
+                "Top k columns": new_top_k,
+            }
+            self._write_json(recommendations)
+            self.get_heatmap()
+            return
+
     def get_clusters(self):
         words = self.rec_table_df["Column"].to_numpy()
         lev_similarity = -1 * np.array(
@@ -136,19 +185,52 @@ def get_clusters(self):
         )
         affprop.fit(lev_similarity)
 
-        print(f"Number of clusters: {np.unique(affprop.labels_).shape[0]}\n")
+        logger.debug(f"Number of clusters: {np.unique(affprop.labels_).shape[0]}\n")
         cluster_names = []
         clusters = {}
         for cluster_id in np.unique(affprop.labels_):
             exemplar = words[affprop.cluster_centers_indices_[cluster_id]]
             cluster = np.unique(words[np.nonzero(affprop.labels_ == cluster_id)])
             cluster_str = ", ".join(cluster)
-            print(" - *%s:* %s" % (exemplar, cluster_str))
+            logger.debug(" - *%s:* %s" % (exemplar, cluster_str))
             cluster_names.append(exemplar)
             clusters[exemplar] = cluster
         self.clusters = clusters
 
-    def _plot_heatmap(self, clusters=[], subschemas=[], threshold=0.5):
+    def _plot_heatmap_base(self, heatmap_rec_list):
+        single = alt.selection_point(name="single")
+        base = (
+            alt.Chart(heatmap_rec_list)
+            .mark_rect(size=100)
+            .encode(
+                y=alt.X("Column:O", sort=None),
+                x=alt.X(f"Recommendation:O", sort=None),
+                color=alt.condition(single, "Value:Q", alt.value("lightgray")),
+                # color="Value:Q",
+                tooltip=[
+                    alt.Tooltip("Column", title="Column"),
+                    alt.Tooltip("Recommendation", title="Recommendation"),
+                    alt.Tooltip("Value", title="Value"),
+                ],
+            )
+            .add_params(single)
+        )
+        return pn.pane.Vega(base)
+
+    def _plot_selected_row(self, heatmap_rec_list, selection):
+        if not selection:
+            return "## No selection"
+        selected_row = heatmap_rec_list.iloc[selection]
+        column = selected_row["Column"].values[0]
+        rec = selected_row["Recommendation"].values[0]
+        # value = selected_row["Value"]
+        # self._accept_match(column, rec)
+        self.selected_row = selected_row
+        return pn.widgets.DataFrame(selected_row)
+
+    def _plot_pane(
+        self, clusters=[], subschemas=[], threshold=0.5, acc_click=0, rej_click=0
+    ):
         heatmap_rec_list = self.rec_list_df[self.rec_list_df["Value"] >= threshold]
         if clusters:
             clustered_cols = []
@@ -165,21 +247,15 @@ def _plot_heatmap(self, clusters=[], subschemas=[], threshold=0.5):
                 heatmap_rec_list["Recommendation"].isin(subschema_rec_cols)
             ]
 
-        base = (
-            alt.Chart(heatmap_rec_list)
-            .mark_rect()
-            .encode(
-                y=alt.X("Column:O", sort=None),
-                x=alt.X(f"Recommendation:O", sort=None),
-                color="Value:Q",
-                tooltip=[
-                    alt.Tooltip("Column", title="Column"),
-                    alt.Tooltip("Recommendation", title="Recommendation"),
-                    alt.Tooltip("Value", title="Value"),
-                ],
-            )
+        heatmap_pane = self._plot_heatmap_base(heatmap_rec_list)
+        return pn.Column(
+            heatmap_pane,
+            pn.bind(
+                self._plot_selected_row,
+                heatmap_rec_list,
+                heatmap_pane.selection.param.single,
+            ),
         )
-        return pn.pane.Vega(base)
 
     def plot_heatmap(self):
         select_cluster = pn.widgets.MultiChoice(
@@ -192,15 +268,35 @@ def plot_heatmap(self):
             name="Threshold", start=0, end=1.0, step=0.01, value=0.5, width=220
         )
 
+        acc_button = pn.widgets.Button(name="Accept Match", button_type="success")
+
+        rej_button = pn.widgets.Button(name="Decline Match", button_type="danger")
+
+        def on_click_accept_match(event):
+            self._accept_match()
+
+        def on_click_reject_match(event):
+            self._reject_match()
+
+        acc_button.on_click(on_click_accept_match)
+        rej_button.on_click(on_click_reject_match)
+
         heatmap_bind = pn.bind(
-            self._plot_heatmap, select_cluster, select_rec_groups, thresh_slider
+            self._plot_pane,
+            select_cluster,
+            select_rec_groups,
+            thresh_slider,
+            acc_button.param.clicks,
+            rej_button.param.clicks,
         )
 
         column_left = pn.Column(
             "# Column",
             select_cluster,
             select_rec_groups,
             thresh_slider,
+            acc_button,
+            rej_button,
             styles=dict(background="WhiteSmoke"),
         )