class sentiment:
host = "localhost"
port = 8086
user = None
password = None
dbname = "mydb2"
db = DBHandler(host, port, user, password, dbname)
dl = DataLoader()
def get_sentiment(self, ticker, company, startDate):
db = self.db
dl = self.dl
startTime = datetime.datetime.strptime(startDate, "%Y-%m-%d")
currentTime = datetime.datetime.today().strftime("%Y-%m-%d")
currentTime = datetime.datetime.strptime(currentTime, "%Y-%m-%d")
days = currentTime - startTime
days = days.days
output = db.QueryDataFromTable("select * from " + ticker +
" where time > now() - " + str(days) +
"d")
df = dl.db_to_df(output)
for i in range(len(df)):
time = df.iloc[[i]].index[0]
if time < datetime.datetime(2006, 6, 30):
avg_polarity = 0.0
avg_subjectivity = 0.0
time = str(time)
time = time.replace(" 00:00:00", "")
else:
time = str(time)
time = time.replace(" 00:00:00", "")
time2 = datetime.datetime.strptime(time, "%Y-%m-%d")
next_day = str(time2 + datetime.timedelta(days=1))
next_day = next_day.replace(" 00:00:00", "")
tweetCriteria = (got.manager.TweetCriteria().setQuerySearch(
company).setSince(time).setUntil(next_day).setMaxTweets(0))
tweets = got.manager.TweetManager.getTweets(tweetCriteria)
if len(tweets) == 0:
avg_polarity = 0.0
avg_subjectivity = 0.0
else:
data = dict()
j = 0
for tweet in tweets:
if "http" not in tweet.text:
analysis = TextBlob(tweet.text)
data[j] = {
"polarity": analysis.sentiment.polarity,
"subjectivity":
analysis.sentiment.subjectivity,
}
j += 1
polarity_total = 0
subjectivity_total = 0
invalid_data_counter = 0
for index in data:
if (data[index]["polarity"] == 0
or data[index]["subjectivity"] == 0):
invalid_data_counter += 1
if invalid_data_counter == len(data):
avg_polarity = 0.0
avg_subjectivity = 0.0
else:
polarity_total = polarity_total + data[index][
"polarity"]
subjectivity_total = (subjectivity_total +
data[index]["subjectivity"])
if len(data) - invalid_data_counter == 0:
avg_polarity = 0.0
avg_subjectivity = 0.0
else:
avg_polarity = polarity_total / (len(data) -
invalid_data_counter)
avg_subjectivity = subjectivity_total / (
len(data) - invalid_data_counter)
output = db.GetRowFromTable(ticker, time)
close = float(output[0][1])
predict = float(output[0][2])
data = [[time, close, predict, avg_polarity, avg_subjectivity]]
df2 = pd.DataFrame(
data,
columns=[
"Date", "Close", "Prediction", "Polarity", "Subjectivity"
],
)
df2["Date"] = pd.to_datetime(df2.Date, format="%Y-%m-%d")
df2.index = df2["Date"]
df2.drop(["Date"], inplace=True, axis=1)
print(df2)
db.DataFrameToDB(df2, ticker)
from alpha_vantage.timeseries import TimeSeries
from dbhelper.db_handler import DBHandler
from time import mktime
import GetOldTweets3 as got
from textblob import TextBlob
from pandas.io.json import json_normalize
host = "localhost"
port = 8086
user = None
password = None
dbname = "mydb"
db = DBHandler(host, port, user, password, dbname)
dl = DataLoader()
status = False
i = 1
ticker = "JPM"
output = db.QueryDataFromTable("select * from " + ticker)
df = dl.db_to_df(output)
print(df)
# time = df.tail(1).index[0]
#!/usr/bin/env python3
from core.get_data import DataLoader
import pandas as pd
import json
import datetime
from alpha_vantage.timeseries import TimeSeries
from dbhelper.db_handler import DBHandler
from time import mktime
import GetOldTweets3 as got
from textblob import TextBlob
dl = DataLoader()
company = "American Airlines"
time = "2012-09-06"
next_day = "2012-09-07"
tweetCriteria = (
got.manager.TweetCriteria()
.setQuerySearch(company)
.setSince(time)
.setUntil(next_day)
.setMaxTweets(1)
)
tweets = got.manager.TweetManager.getTweets(tweetCriteria)
print(tweets)
# data = dict()
#!/usr/bin/env python3 from core.get_data import DataLoader from core.data_processing import DataProcessing from alpha_vantage.timeseries import TimeSeries from core.model import Model from dbhelper.db_handler import DBHandler import pandas as pd import json dl = DataLoader() dp = DataProcessing() host = "localhost" port = 8086 user = None password = None dbname = "mydb2" db = DBHandler(host, port, user, password, dbname) company = "AAL" df = dl.csv_to_df(company + ".csv") print(df) db.DataFrameToDB(df, "testing")
#!/usr/bin/env python3
from core.get_data import DataLoader
from core.model import Model
import pandas as pd
import json
import datetime
from alpha_vantage.timeseries import TimeSeries
from dbhelper.db_handler import DBHandler
from keras import backend as keras
from time import mktime
import GetOldTweets3 as got
from textblob import TextBlob
dl = DataLoader()
md = Model(dl.scaler)
host = "localhost"
port = 8086
user = None
password = None
dbname = "mydb"
db = DBHandler(host, port, user, password, dbname)
df = dl.get_historical_data("AAL")
x_train, y_train = dl.get_train_data(df)
model = md.create_model(x_train, y_train)
#!/usr/bin/env python3
from core.get_data import DataLoader
from core.data_processing import DataProcessing
from alpha_vantage.timeseries import TimeSeries
from core.model import Model
from dbhelper.db_handler import DBHandler
import pandas as pd
import json
dl = DataLoader()
dp = DataProcessing()
host = "localhost"
port = 8086
user = None
password = None
dbname = "mydb2"
db = DBHandler(host, port, user, password, dbname)
output = db.QueryDataFromTable("select * from AAL")
new_output = dl.db_to_df(output)
print(new_output)
dl.df_to_csv(new_output, "AAL.csv")