
from pytrends.request import TrendReq
import datetime as dt
import matplotlib.pyplot as plt
plt.rcParams['figure.figsize'] = [30, 30]
%matplotlib inline  
import numpy as np
import pandas as pd
from sklearn.linear_model import LinearRegression
from sklearn.metrics import mean_squared_error, r2_score
import seaborn as sns


pytrends = TrendReq(hl='en-US', tz=360)  # trend request
kw_list = ["\"opioid abuser\" + \"opioid abusers\" + \"opioid abuse\"", "\"opioid user\" + \"opioid users\" + \"opioid use\"" ]  # keyword list
pytrends.build_payload(kw_list, cat=0, timeframe='2004-01-01 2021-12-31', geo='US', gprop='')  # build pytrends data
trendinfo = pd.DataFrame(pytrends.interest_over_time())  # interest over time, dataframe format


# SHOWS ALL ROWS
#pd.set_option('display.max_rows',trendinfo.shape[0]+1) 

# SHOWS HEAD (first 5 rows in dataframe)
trendinfo.head()


indices = trendinfo.index  # in datetime format

x = np.array(trendinfo['"opioid abuser" + "opioid abusers" + "opioid abuse"']).reshape((-1, 1)) # opioid use array
y = np.array(trendinfo['"opioid user" + "opioid users" + "opioid use"']).reshape((-1, 1)) # opioid abuse array

fig, ax = plt.subplots()
ax.scatter(indices, y, color='g', s=1,label='\"opioid use\" OR \"opioid user\" OR \"opioid users\"')

ax.scatter(indices, x, color='r',s=1, label='\"opioid abuse\" OR \"opioid abuser\" OR \"opioid abusers\"')

plt.legend(bbox_to_anchor=(.08, 1), loc='upper left', borderaxespad=0.)
plt.ylabel('Search Interest')
plt.xlabel('Year')




plt.show()
fig.savefig("opioid_use___opioid_abuse.png", dpi=300)


trendinfo_filtered = trendinfo[trendinfo["\"opioid abuser\" + \"opioid abusers\" + \"opioid abuse\""] > 0]
trendinfo_filtered = trendinfo_filtered[trendinfo_filtered["\"opioid user\" + \"opioid users\" + \"opioid use\""] > 0]


trendinfo_filtered


indices = trendinfo_filtered.index  # in datetime format

x = np.array(trendinfo_filtered['"opioid abuser" + "opioid abusers" + "opioid abuse"']).reshape((-1, 1)) # opioid use array
y = np.array(trendinfo_filtered['"opioid user" + "opioid users" + "opioid use"']).reshape((-1, 1)) # opioid abuse array

fig, ax = plt.subplots()

ax.scatter(indices, y, color='g', s=1,label='\"opioid use\" OR \"opioid user\" OR \"opioid users\"')

ax.scatter(indices, x, color='r', s=1, label='\"opioid abuse\" OR \"opioid abuser\" OR \"opioid abuser\"')

plt.legend(bbox_to_anchor=(.08, 1), loc='upper left', borderaxespad=0.)
plt.ylabel('Search Interest')
plt.xlabel('Year')

plt.show()
fig.savefig("opioid_use___opioid_abuse_zero_removed.png", dpi=300)


fig, ax = plt.subplots()

ax.scatter(indices, y, color='g', s=1,label='\"opioid use\" OR \"opioid user\" OR \"opioid users\"')

ax.scatter(indices, x, color='r', s=1, label='\"opioid abuse\" OR \"opioid abuser\" OR \"opioid abusers\"')

plt.legend(bbox_to_anchor=(.07, 1), loc='upper left', borderaxespad=0.)
plt.ylabel('Search Interest')
plt.xlabel('Year')

plt.show()
fig.savefig("opioid_use___opioid_abuse_zero_removed.png", dpi=300)

	"opioid abuser" + "opioid abusers" + "opioid abuse"	"opioid user" + "opioid users" + "opioid use"	isPartial
date
2004-01-01	0	0	False
2004-02-01	0	0	False
2004-03-01	0	0	False
2004-04-01	0	0	False
2004-05-01	0	0	False

	"opioid abuser" + "opioid abusers" + "opioid abuse"	"opioid user" + "opioid users" + "opioid use"	isPartial
date
2007-12-01	9	9	False
2008-07-01	15	8	False
2009-01-01	7	7	False
2009-03-01	6	6	False
2009-10-01	6	6	False
...	...	...	...
2021-08-01	12	55	False
2021-09-01	11	72	False
2021-10-01	20	73	False
2021-11-01	18	76	False
2021-12-01	8	46	False

OPIOD_USE___OPOIOD_ABUSE_25JUL2021¶

Remove rows with zeros¶