Update trends.py

Analytic_Web_Apps/SocialMedia_Twitter/trends.py

@@ -1,26 +1,117 @@
 import plotly.express as px
 import dash_bootstrap_components as dbc
 import dash
+import dash_table
 import dash_html_components as html
 import dash_core_components as dcc
 from dash.dependencies import Output, Input
 import pandas as pd
 import twitter  # pip install python-twitter
 from app import app, api
+import nltk
+from wordcloud import WordCloud
+# nltk.download('stopwords') # only downlaod once
+# Resource: https://python-twitter.readthedocs.io/en/latest/twitter.html
+
+# Create stopword list for wordcloud analysis later:
+stopwords = nltk.corpus.stopwords.words('english')
+stopwords = set(stopwords)
+stopwords.update(["https", "plotlygraphs"])
+
+def f(row):
+    return "[{0}]({0})".format(row["url"])
 
 # layout of second (trends) tab ******************************************
-trends_layout = html.Div([
+trends_layout = dbc.Container([
     dbc.Row([
         dbc.Col([
-            html.H2("Every 10 seconds see the most trending tweets and their analysis")
+            html.H2("Most trending topics & Word Cloud Analysis of tweets")
         ], width=12)
     ]),
     dbc.Row([
         dbc.Col([
-            dcc.Graph(id='scatter', figure={})
+            html.Div(id='table-div', children="")
         ], width=6),
         dbc.Col([
-            dcc.Graph(id='scatter2', figure={})
-        ], width=6)
-    ])
-])
+            html.Div(id='figure-div', children="")
+        ], width=6),
+    ], className="mt-3",),
+    dcc.Interval(id='timer', interval=1000*300, n_intervals=0)
+])
+
+
+# pull trending tweets and create the table ******************************
+@app.callback(
+    Output(component_id="table-div", component_property="children"),
+    Input(component_id="timer", component_property="n_intervals"),
+)
+def display_trend(timer):
+    trnd_name, trnd_vol, trnd_url = [], [], []
+    trends = api.GetTrendsCurrent()
+    for trend in trends:
+        print(trend)
+        if trend.tweet_volume:
+            trnd_name.append(trend.name)
+            trnd_vol.append(trend.tweet_volume)
+            trnd_url.append(trend.url)
+    d = {
+        "trending": trnd_name,
+        "url": trnd_url,
+        "volume": trnd_vol,
+    }
+    df = pd.DataFrame(d)
+    # aply function so you can insert url link inside Dash DataTable
+    df["url"] = df.apply(f, axis=1)
+    print(df.head())
+
+    return dash_table.DataTable(
+        id='datatable-trends',
+        columns=[
+            {"name": i, "id": i}
+            if i == "trending" or i == "volume"
+            else {"name": i, "id": i, 'type': 'text', "presentation":"markdown"}
+            for i in df.columns
+        ],
+        data=df.to_dict('records'),
+        markdown_options=dict(html=True, link_target='_blank'),
+        page_action='native',
+        page_size=6,
+        style_cell={
+            'whiteSpace': 'normal',
+            'height': 'auto',
+            'overflow': 'hidden',
+            'minWidth': '50px', 'width': '80px', 'maxWidth': '120px',
+        },
+    )
+
+
+# pull liked tweets by user **********************************************
+@app.callback(
+    Output(component_id="figure-div", component_property="children"),
+    Input(component_id="timer", component_property="n_intervals"),
+)
+def display_trend(timer):
+    liked_twt = []
+    # get tweets metadata that were liked by @Plotlygraphs
+    x = api.GetFavorites(screen_name='plotlygraphs', count=100, return_json=False)
+    # extract the tweet text, from the metadata, into a list
+    for status in x:
+        liked_twt.append(status.text)
+
+    # join all tweet text into one string
+    alltweets = " ".join(tweet for tweet in liked_twt)
+
+    # generate wordcloud from all tweets
+    my_wordcloud = WordCloud(
+        stopwords=stopwords,
+        background_color='white',
+        height=275
+    ).generate(alltweets)
+
+    # visualize wordcloud inside plotly figure
+    fig = px.imshow(my_wordcloud, template='ggplot2')
+    fig.update_layout(margin=dict(l=20, r=20, t=30, b=20))
+    fig.update_xaxes(visible=False)
+    fig.update_yaxes(visible=False)
+
+    return dcc.Graph(figure=fig, config={"displayModeBar":False})