From 32c4916e1c90a052a8db57bea78c13f61add6d1a Mon Sep 17 00:00:00 2001
From: "intern-william@nodeflux-67" <justin@nodeflux.io>
Date: Mon, 27 Jun 2022 09:43:52 +0000
Subject: [PATCH] update embedding GUI

---
 app_dash.py        |  3 +-
 pages/embedding.py | 93 ++++++++++++++++++++++++++++++++--------------
 requirements.txt   |  1 +
 3 files changed, 68 insertions(+), 29 deletions(-)

diff --git a/app_dash.py b/app_dash.py
index 8a07d4c..7eef4a4 100644
--- a/app_dash.py
+++ b/app_dash.py
@@ -1,7 +1,8 @@
 import dash
+import dash_bootstrap_components as dbc
 import config
 
-app = dash.Dash(__name__, use_pages=True)
+app = dash.Dash(__name__, use_pages=True, external_stylesheets=[dbc.themes.BOOTSTRAP])
 app.layout = dash.html.Div([
 	dash.page_container
 ])
diff --git a/pages/embedding.py b/pages/embedding.py
index e21aae8..1bd8aca 100644
--- a/pages/embedding.py
+++ b/pages/embedding.py
@@ -1,7 +1,9 @@
 import dash
+import dash_bootstrap_components as dbc
 import pandas as pd
 import plotly.express as px
 import plotly.graph_objects as go
+import math
 import requests
 import os
 import config
@@ -11,9 +13,7 @@ def create_figure(df, min_uniqueness, max_uniqueness, min_sqrt_area, max_sqrt_ar
     fig = px.scatter(df[mask], x='embeddings_x', y='embeddings_y', color='label', size='sqrt_area', custom_data=['id'], hover_data=['uniqueness'])
     figure = go.FigureWidget(fig)
     figure.update_layout(
-        autosize=True,
-        width=1080,
-        height=566,
+        dragmode='lasso'
     )
     return figure
 
@@ -27,33 +27,70 @@ def main(name):
     global df
     df = pd.read_pickle(file_path)
     uniqueness_range = (0, 1)
-    sqrt_area_range = (0, df['sqrt_area'].max())
+    sqrt_area_range = (0, math.ceil(df['sqrt_area'].max()))
 
     figure = create_figure(df, uniqueness_range[0], uniqueness_range[1], sqrt_area_range[0], sqrt_area_range[1])
 
-    return dash.html.Div(children=[
-        dash.html.H3('Embedding Visualization'),
-        dash.dcc.Graph(
-            id='graph',
-            figure=figure
-        ),
-        dash.html.P(f'{name}', id='name'),
-        dash.html.P(id='num_sample'),
-        dash.html.P('Filter by uniqueness:'),
-        dash.dcc.RangeSlider(
-            id='uniqueness-slider',
-            min=0, max=1, step=0.001,
-            marks={0: '0', 1: '1'},
-            value=[0, 1]
-        ),
-        dash.html.P('Filter by sqrt area:'),
-        dash.dcc.RangeSlider(
-            id='sqrt-area-slider',
-            min=0, max=df['sqrt_area'].max(), step=df['sqrt_area'].max()/1000,
-            marks={0: '0', df['sqrt_area'].max(): f'{df["sqrt_area"].max()}'},
-            value=[0, df['sqrt_area'].max()]
-        ),
-    ])        
+    return dbc.Container(
+        children = [
+            dash.html.H1('Embedding Visualization', style = {'text-align':'center'}),
+            dbc.Container(
+                children = [
+                    dash.html.P(
+                        '''
+                        Each node in this graph represents a bounding box in the dataset.
+                        Use 'box select' or 'lasso select' to select nodes. Double click on a label to excluded the other label.
+                        The slider will filter nodes that do not meet the criteria. Note that the slider don't update the selection.
+                        Selected nodes on this graph can be previewed on Fiftyone through the button on the bottom of this page.
+                        '''
+                    ),
+                ],
+                className = 'm-3'
+            ),
+            dash.html.Div(
+                dash.dcc.Graph(
+                    id='graph',
+                    figure=figure,
+                    responsive=True,
+                    style={'width' : '100%', 'height': '100%'}
+                ),
+                style={'width' : '100%', 'height': '60%'}
+            ),
+            dbc.Container(
+                children = [
+                    dash.html.P(f'{name}', id='name'),
+                    dash.html.P(id='num_sample'),
+                ],
+                className = 'm-3'
+            ),
+            dbc.Container(
+                children = [
+                    dash.html.B('Filter by Uniqueness'),
+                    dash.dcc.RangeSlider(
+                        id='uniqueness-slider',
+                        min=uniqueness_range[0], max=uniqueness_range[1], step=(uniqueness_range[1]-uniqueness_range[0])/1000,
+                        marks={uniqueness_range[0]: uniqueness_range[0], uniqueness_range[1]: uniqueness_range[1]},
+                        value=[uniqueness_range[0], uniqueness_range[1]]
+                    )
+                ],
+                className = 'm-3'
+            ),
+            dbc.Container(
+                children = [
+                    dash.html.B('Filter by Sqrt Area'),
+                    dash.dcc.RangeSlider(
+                        id='sqrt-area-slider',
+                        min=sqrt_area_range[0], max=sqrt_area_range[1], step=(sqrt_area_range[1]-sqrt_area_range[0])/1000,
+                        marks={sqrt_area_range[0]: sqrt_area_range[0], sqrt_area_range[1]: sqrt_area_range[1]},
+                        value=[sqrt_area_range[0], sqrt_area_range[1]]
+                    ),
+                ],
+                className = 'm-3'
+            ),
+            dbc.Button('Open Fiftyone', color='light', className='m-3', href=f'{config.url}:{config.port["flask"]}/fiftyone/{name}', external_link=True, target='_blank'),
+        ],
+        className = 'p-5 vh-100 vw-100'
+    )
 
 @dash.callback(
     dash.Output(component_id='graph', component_property='figure'), 
@@ -88,7 +125,7 @@ def update(name, input_value):
         }
     )
     return f'Number of patches: {len(ids)}'
-    
+
 dash.register_page(__name__, path_template="/embedding/<name>")
 
 def layout(name=None):
diff --git a/requirements.txt b/requirements.txt
index fbe2a4e..012ec22 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,5 +1,6 @@
 Flask==2.1.2
 dash==2.5.1
+dash-bootstrap-components==1.1.0
 plotly==5.9.0
 ipywidgets==7.7.1
 torch==1.11.0