app.py

from flask import Flask, render_template, request
from numpy import array, mean
import pickle
from datetime import datetime, timedelta
from pymongo import MongoClient
import plotly
import plotly.express as px
import pandas as pd
import json
import numpy as np
from google.cloud import translate_v2 as translate


try:
	client = MongoClient("mongodb+srv://sjhbluhm:123password!@cluster0-o0tfo.mongodb.net/test?retryWrites=true&w=majority")
	client.server_info()
	db = client["moodring"]
	collection = db["moodring"]
	print("connected to Mongodb server")
except:
	print("connection failure")


w2v_model = None
sent_model = None
translate_client = None


app = Flask(__name__)
if __name__ == '__main__':
    app.run()
    app.config["TEMPLATES_AUTO_RELOAD"]



@app.route("/")
def hello():
    bar = create_plot()
    
	#create array of previous entries newest to oldest
    arr_entries = []
    global collection

    for result in collection.find({}).sort("dtstamp",-1):
        i = result["dtstamp"]
        j = result["text"]
        k = result["sentiment"]
        arr_entries.append([i, j, k])

    #find the average of past day's sentiment
    sum = entries = 0
    datetimestamp=datetime.utcnow()
    results = collection.find({"day":datetimestamp.strftime("%d "),"month":datetimestamp.strftime("%m "), "year":datetimestamp.strftime("%Y ")})
    for result in results:
        sum += result["sentiment"]
        entries += 1
    if entries > 0:
    	todaysentiment = round(sum/entries,2)

    else:
        todaysentiment = 0
    return render_template('index.html', plot=bar, arr_entries=arr_entries, index="active",entries="inactive", todaysentiment=todaysentiment)


@app.route("/add_entry", methods=["GET", "POST"])
# default goal_display is current time, at EST. takes in form input if posted
def add_entry():
    if request.method == "GET":
        return render_template("addentry.html")
    else:
        global sent_model
        global w2v_model
        global translate_client
        if not sent_model:
            sent_model = pickle.load(open('ml_code/model.sav', 'rb'))
        if not w2v_model:
            w2v_model = pickle.load(open('ml_code/vectors.sav', 'rb'))
        if not translate_client:
            translate_client = translate.Client()
        journal = request.form.get("journal")
        sentiment=get_sentiment(journal)
        datetimestamp=datetime.utcnow()
        entry = {"dtstamp":datetimestamp, "text":journal, "sentiment":sentiment, "hour":datetimestamp.strftime("%H "),
                 "day":datetimestamp.strftime("%d "), "month":datetimestamp.strftime("%m "),
                 "year":datetimestamp.strftime("%Y ")}
        global collection
        collection.insert_one(entry)
        return render_template("addentry.html")


def get_sentiment(entry):
    global w2v_model
    global sent_model
    global translate_client
    response = translate_client.translate(entry, target_language='en')
    translation = response['translatedText']
    word_list = translation.split()
    sentiment_list = []
    for word in word_list:
        lowercase_word = word.lower()
        if lowercase_word in w2v_model:
            sentiment_list.append(w2v_model[lowercase_word])
    if len(sentiment_list) > 0:
    	vector_array = array(sentiment_list)
    	avg_sent = mean(vector_array, axis=0).reshape(1, -1)
    	result = sent_model.predict(avg_sent)[0]
    else:
    	result = 0
    return int(result)


def create_plot():
    global collection
    results = collection.find({})
    journals = []
    tags = []
    dates = []
    for r in results:
        journals.append(r["text"])
        dates.append(r['day'])
        tags.append(r['sentiment'])
    ids = range(1,len(journals)+1)
    sample_df = df = pd.DataFrame(
        {'id': ids,
        'review': journals,
        'tag': tags,
        'date': dates
        }
    )

    hraverage = []
    now = datetime.utcnow()
    for i in range(24):
        sum = 0
        entries = 0
        then = now - timedelta(hours=i)
        results = collection.find({"hour":then.strftime("%H "),"day":then.strftime("%d "),
                                   "month":then.strftime("%m "), "year":then.strftime("%Y ")})
        for result in results:
            sum += result["sentiment"]
            entries += 1
        if entries > 0:
            avrg = sum/entries
        else:
            avrg = 0
        hraverage.append(avrg)
    avg_ids = range(24,0,-1)
    sample_df2 = df2 = pd.DataFrame(
        {'id': avg_ids,
        'average':hraverage
        }
    )
    # graph option 2: polar barplot w/ binary height, color represents string length
    fig2 = px.bar_polar(
        sample_df2,
        r=sample_df2['average'].add(1),
        theta=np.arange(0, 360, 360/24),
        color=sample_df2['average'],
        color_continuous_scale=["#352961", "#be9fe1", "#fdd365"],
        range_color=[-1, 1],
        )
    fig2.update_layout(
        title='World Mood',
        plot_bgcolor='rgb(10,10,10)',
        
        polar = dict(
            radialaxis = dict(range=[0.5,2.2], showticklabels=False, ticks=''),
            angularaxis = dict(showticklabels=False, ticks='')
        ),
        coloraxis=dict(
            colorbar=dict(
                title='Average Hourly Sentiment',
                tickmode='array',
                tickvals=[-1, 0, 1],
                ticktext=['-1: negative', '0: neutral', '1: positive']))
    )

    graphJSON = json.dumps(fig2, cls=plotly.utils.PlotlyJSONEncoder)
    return graphJSON