iRobot Stock Analysis
In this Jupyter Notebook I have done some exploratory data analysis of iRobot stock.
import quandl
import pandas as pd
import numpy as np
import pandas_datareader as pdr
#from datetime import datetime
import datetime
from __future__ import division
import math
from sklearn import preprocessing, cross_validation, svm
from sklearn.linear_model import LinearRegression
import pickle
import matplotlib.pyplot as plt
import seaborn as sns
sns.set_style('whitegrid')
%matplotlib inline
#Set times, -1 = yearg ago from now
end = datetime.now()
print (end)
start = datetime(end.year -1, end.month, end.day)
print (start)
2018-06-07 22:12:03.122418
2017-06-07 00:00:00
#Import data from yahoo and assign it to IRBT
IRBT = pdr.DataReader('IRBT', 'yahoo', start, end)
IRBT.info()
<class 'pandas.core.frame.DataFrame'>
DatetimeIndex: 253 entries, 2017-06-07 to 2018-06-07
Data columns (total 6 columns):
Open 253 non-null float64
High 253 non-null float64
Low 253 non-null float64
Close 253 non-null float64
Adj Close 253 non-null float64
Volume 253 non-null int64
dtypes: float64(5), int64(1)
memory usage: 13.8 KB
#View historical trend of the closing price Adj Close column
IRBT['Adj Close'].plot(legend=True, figsize=(10,4))
<matplotlib.axes._subplots.AxesSubplot at 0x11efda048>
IRBT.head()
| Open | High | Low | Close | Adj Close | Volume | |
|---|---|---|---|---|---|---|
| Date | ||||||
| 2017-06-07 | 96.860001 | 99.309998 | 96.080002 | 99.230003 | 99.230003 | 687500 |
| 2017-06-08 | 98.980003 | 100.480003 | 97.820000 | 99.930000 | 99.930000 | 633700 |
| 2017-06-09 | 100.419998 | 100.949997 | 93.129997 | 95.989998 | 95.989998 | 1135700 |
| 2017-06-12 | 95.989998 | 98.760002 | 91.510002 | 98.650002 | 98.650002 | 1133200 |
| 2017-06-13 | 99.430000 | 101.540001 | 98.500000 | 100.279999 | 100.279999 | 801200 |
IRBT.tail()
| Open | High | Low | Close | Adj Close | Volume | |
|---|---|---|---|---|---|---|
| Date | ||||||
| 2018-06-01 | 62.820000 | 64.349998 | 62.820000 | 63.730000 | 63.730000 | 619600 |
| 2018-06-04 | 64.099998 | 65.639999 | 64.099998 | 65.360001 | 65.360001 | 659500 |
| 2018-06-05 | 65.290001 | 66.080002 | 65.000000 | 65.919998 | 65.919998 | 510500 |
| 2018-06-06 | 66.220001 | 69.250000 | 65.949997 | 68.349998 | 68.349998 | 933200 |
| 2018-06-07 | 68.160004 | 69.290001 | 67.290001 | 69.040001 | 69.040001 | 625500 |
df = IRBT[['Open','High','Low','Close', 'Volume']]
df['HL_PCT'] = (df['High']-df['Close'])/ df['Close'] * 100.0
df['PCT_Change'] = (df['Close']-df['Open'])/ df['Open'] * 100.0
df = df [['Close','HL_PCT','PCT_Change','Volume']]
df.tail()
| Close | HL_PCT | PCT_Change | Volume | |
|---|---|---|---|---|
| Date | ||||
| 2018-06-01 | 63.730000 | 0.972851 | 1.448583 | 619600 |
| 2018-06-04 | 65.360001 | 0.428394 | 1.965683 | 659500 |
| 2018-06-05 | 65.919998 | 0.242725 | 0.964921 | 510500 |
| 2018-06-06 | 68.349998 | 1.316755 | 3.216546 | 933200 |
| 2018-06-07 | 69.040001 | 0.362109 | 1.291075 | 625500 |
#Define a label -- Adj Price in the future
forecast_col = 'Close'
df.fillna(-99999, inplace=True)
forecast_out = int(math.ceil(0.01*len(df)))
df['Label'] = df[forecast_col].shift(-forecast_out)
df.dropna(inplace=True)
df.head()
| Close | HL_PCT | PCT_Change | Volume | Label | |
|---|---|---|---|---|---|
| Date | |||||
| 2017-06-07 | 99.230003 | 0.080616 | 2.446833 | 687500 | 98.650002 |
| 2017-06-08 | 99.930000 | 0.550388 | 0.959787 | 633700 | 100.279999 |
| 2017-06-09 | 95.989998 | 5.167204 | -4.411472 | 1135700 | 101.989998 |
| 2017-06-12 | 98.650002 | 0.111505 | 2.771126 | 1133200 | 97.970001 |
| 2017-06-13 | 100.279999 | 1.256484 | 0.854872 | 801200 | 96.269997 |
df[['Close','Label']].plot(subplots=False,legend=True, figsize=(10,4))
<matplotlib.axes._subplots.AxesSubplot at 0x121a8f588>
#Define features X and labels y
X = np.array(df.drop(['Label'],1))
y = np.array(df['Label'])
#Scale X
X = preprocessing.scale(X)
y = np.array(df['Label'])
print(len(X),len(y))
250 250
#Create training and testing sets
X_train, X_test, y_train, y_test = cross_validation.train_test_split(X, y, test_size=0.2)
#Define a classifier and test it .fit=train / .score=test
classf = LinearRegression()
classf.fit(X_train, y_train)
accuracy = classf.score(X_test, y_test)
print(forecast_out)
print(accuracy)
3
0.849147136972
#Using the Support Vector Machine algo
classf = svm.SVR()
classf.fit(X_train, y_train)
accuracy = classf.score(X_test, y_test)
print(forecast_out)
print(accuracy)
3
0.708794197356
#Forecasting into the future...
df['Label'] = df[forecast_col].shift(-forecast_out)
X = np.array(df.drop(['Label'],1))
X = preprocessing.scale(X)
X_lately = X[-forecast_out:]
X = X[:-forecast_out]
df.dropna(inplace=True)
y = np.array(df['Label'])
#Create training and testing sets
X_train, X_test, y_train, y_test = cross_validation.train_test_split(X, y, test_size=0.2)
#Define a classifier and test it .fit=train / .score=test
classf = LinearRegression()
classf.fit(X_train, y_train)
accuracy = classf.score(X_test, y_test)
print(forecast_out)
print(accuracy)
3
0.925348943641
forecast_set = classf.predict(X_lately)
print (forecast_set, accuracy, forecast_out)
[ 62.91718632 63.80245711 65.64647982] 0.925348943641 3
df['Forecast']= np.nan
last_date = df.iloc[-1].name
last_unix = last_date.timestamp()
one_day = 86400
next_unix = last_unix + one_day
for i in forecast_set:
next_date = datetime.datetime.fromtimestamp(next_unix)
next_unix += one_day
df.loc[next_date] = [np.nan for _ in range(len(df.columns)-1)]+ [i]
df['Close'].plot()
df['Forecast'].plot()
plt.xlabel('Date')
plt.ylabel('Price')
plt.show()
df[['Close','Forecast']].plot(subplots=False,legend=True, figsize=(10,4))
<matplotlib.axes._subplots.AxesSubplot at 0x121f011d0>
df
| Close | HL_PCT | PCT_Change | Volume | Label | Forecast | |
|---|---|---|---|---|---|---|
| 2017-06-07 | 99.230003 | 0.080616 | 2.446833 | 687500.0 | 98.650002 | NaN |
| 2017-06-08 | 99.930000 | 0.550388 | 0.959787 | 633700.0 | 100.279999 | NaN |
| 2017-06-09 | 95.989998 | 5.167204 | -4.411472 | 1135700.0 | 101.989998 | NaN |
| 2017-06-12 | 98.650002 | 0.111505 | 2.771126 | 1133200.0 | 97.970001 | NaN |
| 2017-06-13 | 100.279999 | 1.256484 | 0.854872 | 801200.0 | 96.269997 | NaN |
| 2017-06-14 | 101.989998 | 2.568882 | 1.644411 | 1283400.0 | 98.779999 | NaN |
| 2017-06-15 | 97.970001 | 3.143822 | -2.225545 | 1657400.0 | 97.730003 | NaN |
| 2017-06-16 | 96.269997 | 2.690354 | -2.442241 | 1599000.0 | 101.080002 | NaN |
| 2017-06-19 | 98.779999 | 0.496050 | 1.877063 | 883400.0 | 100.430000 | NaN |
| 2017-06-20 | 97.730003 | 2.302261 | -1.282825 | 604500.0 | 101.199997 | NaN |
| 2017-06-21 | 101.080002 | 1.226749 | 2.723577 | 780200.0 | 100.980003 | NaN |
| 2017-06-22 | 100.430000 | 1.214778 | -0.554508 | 560200.0 | 90.650002 | NaN |
| 2017-06-23 | 101.199997 | 1.482213 | 0.806851 | 696400.0 | 92.470001 | NaN |
| 2017-06-26 | 100.980003 | 1.871658 | -0.951441 | 392100.0 | 86.779999 | NaN |
| 2017-06-27 | 90.650002 | 10.821839 | -9.675168 | 1906600.0 | 84.139999 | NaN |
| 2017-06-28 | 92.470001 | 1.254457 | 1.581898 | 1359000.0 | 83.540001 | NaN |
| 2017-06-29 | 86.779999 | 6.914036 | -6.426568 | 1299900.0 | 82.180000 | NaN |
| 2017-06-30 | 84.139999 | 3.220821 | -2.412436 | 1125500.0 | 80.540001 | NaN |
| 2017-07-03 | 83.540001 | 1.460379 | 0.288113 | 886300.0 | 83.650002 | NaN |
| 2017-07-05 | 82.180000 | 2.579707 | -1.035647 | 1140600.0 | 84.470001 | NaN |
| 2017-07-06 | 80.540001 | 1.092621 | -0.297101 | 1400600.0 | 83.239998 | NaN |
| 2017-07-07 | 83.650002 | 1.016136 | 4.171854 | 989000.0 | 85.019997 | NaN |
| 2017-07-10 | 84.470001 | 0.627441 | 0.607434 | 776500.0 | 84.620003 | NaN |
| 2017-07-11 | 83.239998 | 2.570878 | -1.257420 | 812600.0 | 83.989998 | NaN |
| 2017-07-12 | 85.019997 | 0.270528 | 1.069893 | 814900.0 | 84.900002 | NaN |
| 2017-07-13 | 84.620003 | 1.217205 | -1.202567 | 527500.0 | 84.750000 | NaN |
| 2017-07-14 | 83.989998 | 1.154901 | -0.896758 | 424200.0 | 86.769997 | NaN |
| 2017-07-17 | 84.900002 | 0.895173 | 1.035348 | 556900.0 | 85.000000 | NaN |
| 2017-07-18 | 84.750000 | 0.436582 | -0.200186 | 554000.0 | 88.470001 | NaN |
| 2017-07-19 | 86.769997 | 0.806735 | 2.070339 | 559200.0 | 90.379997 | NaN |
| ... | ... | ... | ... | ... | ... | ... |
| 2018-04-23 | 59.950001 | 5.054210 | -4.018569 | 2061800.0 | 58.240002 | NaN |
| 2018-04-24 | 59.080002 | 4.807034 | -2.039461 | 2178300.0 | 58.580002 | NaN |
| 2018-04-25 | 57.060001 | 8.499823 | -7.789269 | 2951800.0 | 58.360001 | NaN |
| 2018-04-26 | 58.240002 | 2.970465 | 2.050117 | 1423100.0 | 57.830002 | NaN |
| 2018-04-27 | 58.580002 | 1.502214 | 0.652923 | 1150700.0 | 58.310001 | NaN |
| 2018-04-30 | 58.360001 | 0.822479 | -0.273409 | 616500.0 | 57.180000 | NaN |
| 2018-05-01 | 57.830002 | 0.916478 | -0.344644 | 709300.0 | 60.020000 | NaN |
| 2018-05-02 | 58.310001 | 0.840333 | 1.144838 | 549700.0 | 61.759998 | NaN |
| 2018-05-03 | 57.180000 | 2.168587 | -1.464758 | 654200.0 | 62.270000 | NaN |
| 2018-05-04 | 60.020000 | 0.216598 | 5.520392 | 850600.0 | 62.060001 | NaN |
| 2018-05-07 | 61.759998 | 0.259067 | 2.404240 | 824500.0 | 61.930000 | NaN |
| 2018-05-08 | 62.270000 | 0.112414 | 0.630250 | 497600.0 | 62.360001 | NaN |
| 2018-05-09 | 62.060001 | 0.805672 | -0.369241 | 517900.0 | 62.299999 | NaN |
| 2018-05-10 | 61.930000 | 1.356370 | -0.625802 | 584900.0 | 61.259998 | NaN |
| 2018-05-11 | 62.360001 | 0.384857 | 0.694334 | 485200.0 | 62.310001 | NaN |
| 2018-05-14 | 62.299999 | 1.364371 | -0.272134 | 509600.0 | 62.880001 | NaN |
| 2018-05-15 | 61.259998 | 1.142672 | -0.969937 | 661000.0 | 62.430000 | NaN |
| 2018-05-16 | 62.310001 | 0.866630 | 1.070558 | 462000.0 | 62.889999 | NaN |
| 2018-05-17 | 62.880001 | 0.795165 | 0.930982 | 693100.0 | 63.160000 | NaN |
| 2018-05-18 | 62.430000 | 0.736824 | -0.731434 | 464000.0 | 61.869999 | NaN |
| 2018-05-21 | 62.889999 | 0.079504 | 0.143312 | 633900.0 | 62.389999 | NaN |
| 2018-05-22 | 63.160000 | 2.184929 | -0.158075 | 621600.0 | 62.730000 | NaN |
| 2018-05-23 | 61.869999 | 1.874899 | -1.574932 | 654600.0 | 64.839996 | NaN |
| 2018-05-24 | 62.389999 | 0.112197 | 0.873082 | 434200.0 | 65.300003 | NaN |
| 2018-05-25 | 62.730000 | 0.031883 | 0.593327 | 380900.0 | 62.410000 | NaN |
| 2018-05-29 | 64.839996 | 0.046279 | 3.413070 | 739800.0 | 63.730000 | NaN |
| 2018-05-30 | 65.300003 | 0.459410 | 0.554359 | 644800.0 | 65.360001 | NaN |
| 2018-05-31 | NaN | NaN | NaN | NaN | NaN | 62.917186 |
| 2018-06-01 | NaN | NaN | NaN | NaN | NaN | 63.802457 |
| 2018-06-02 | NaN | NaN | NaN | NaN | NaN | 65.646480 |
250 rows × 6 columns