dennisthiessen/stock-prediction
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Added SMA and improved plotting

Browse Source
This commit is contained in:
Dennis Thiessen
2017-12-15 15:16:22 +01:00
parent 055ada2766
commit 5c08182aad
1 changed files with 20 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

22
bot/core.py
View File

@@ -2,12 +2,15 @@
import time import time
import pandas as pd import pandas as pd
import numpy as np import numpy as np
import datetime as dt
import matplotlib.pyplot as plt import matplotlib.pyplot as plt
import matplotlib.dates as md
import tensorflow as tf import tensorflow as tf
import urllib.request, json import urllib.request, json
from bot.shared_config import * from bot.shared_config import *
def main(): def main():
start_time = time.time() start_time = time.time()
@@ -29,7 +32,7 @@ def main():
dump(green("Retrieved API in {0:.3f}ms sec".format((time.time() - start_time)*100))) dump(green("Retrieved API in {0:.3f}ms sec".format((time.time() - start_time)*100)))
dump(yellow("Initialize Tensorflow")) dump(yellow("Initialize Tensorflow"))
f_horizon = 1 # forecast horizon, one period into the future f_horizon = 10 # forecast horizon, one period into the future
num_periods = 20 # number of periods per vector we are using to predict one period ahead num_periods = 20 # number of periods per vector we are using to predict one period ahead
inputs = 2 # number of vectors submitted inputs = 2 # number of vectors submitted
hidden = 100 # number of neurons we will recursively work through, can be changed to improve accuracy hidden = 100 # number of neurons we will recursively work through, can be changed to improve accuracy
@@ -98,9 +101,20 @@ def main():
actual_prediction = pd.Series(np.concatenate([np.ravel(X_test)[::2], np.ravel(y_pred)])) actual_prediction = pd.Series(np.concatenate([np.ravel(X_test)[::2], np.ravel(y_pred)]))
plt.title("Forecast vs Actual", fontsize=14) plt.title("Forecast vs Actual", fontsize=14)
plt.plot(actual_series, "b-", markersize=10, label="Actual")
# plt.plot(pd.Series(np.ravel(Y_test)), "w*", markersize=10) xfmt = md.DateFormatter('%d.%m %H:%M:%S')
plt.plot(actual_prediction, "r-", markersize=7, label="Forecast") plt.subplots_adjust(bottom=0.2)
plt.xticks(rotation=25)
plt.gca().xaxis.set_major_formatter(xfmt)
dates = [dt.datetime.fromtimestamp(int(ts)) for ts in timestamps]
datenums = md.date2num(dates)
plt.plot(datenums[-40:], actual_series, "b--", linewidth=1.0, label="Actual")
plt.plot(datenums[-40:], actual_prediction, "r--", linewidth=1.0, label="Forecast")
plt.plot(datenums[-40:], actual_series.rolling(window=3, center=False).mean(), "y-", linewidth=2.0, label="Actual MA")
plt.plot(datenums[-40:], actual_prediction.rolling(window=3, center=False).mean(), "g-", linewidth=2.0, label="Predicted MA")
plt.legend(loc="upper left") plt.legend(loc="upper left")
plt.xlabel("Time Periods") plt.xlabel("Time Periods")