def main():
print(" \n NHL Player Predictions\n", "=" * 23)
print("Legend:", "LW = Left Wing,", "RW = Right Wing,", "C = Center,",
"D = Defenseman")
db_data = DataStore()
df = db_data.fetch_all()
df.drop(['id'], 1, inplace=True)
df.drop(['team'], 1, inplace=True)
play_sample = df.sample(7)
df.drop(['name'], 1, inplace=True)
df.drop(play_sample.index)
X = np.array(df.drop(['position'], 1))
y = np.array(df['position'])
X_train, X_test, y_train, y_test = model_selection.train_test_split(
X, y, test_size=0.3)
clf = neighbors.KNeighborsClassifier()
clf.fit(X_train, y_train)
accuracy = clf.score(X_test, y_test)
print("Number of Data Points:", len(df.index), "\n")
print("\nPredicted with Accuracy: ", accuracy * 100, "%\n")
# Prediction Test
names = np.array(play_sample['name'])
positions = np.array(play_sample['position'])
players = dict(zip(names, positions))
play_sample.drop(['name'], 1, inplace=True)
play_sample.drop(['position'], 1, inplace=True)
X_play = np.array(play_sample)
predictions = clf.predict(X_play)
outcome = []
for i in range(len(predictions)):
if predictions[i] == positions[i]:
outcome.append("PASS")
else:
outcome.append("FAIL")
output = pd.DataFrame(data=np.column_stack(
(predictions, positions, outcome)),
index=names,
columns=["Predicted", "Actual", "Outcome"])
print(output)
