YAO/YAO-optimizer
1
0
Fork 0
mirror of https://github.com/newnius/YAO-optimizer.git synced 2025-06-07 07:01:56 +00:00
Code Issues Releases Wiki Activity

update

Browse Source
This commit is contained in:
Newnius 2020-05-02 10:17:38 +08:00
parent edf61d117d
commit 4bd70489c5
2 changed files with 69 additions and 60 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
.idea/workspace.xml
View File

@ -3,7 +3,6 @@
<component name="ChangeListManager"> <component name="ChangeListManager">
<list default="true" id="0aedafd8-e57e-462a-beda-65af0b91f3df" name="Default Changelist" comment=""> <list default="true" id="0aedafd8-e57e-462a-beda-65af0b91f3df" name="Default Changelist" comment="">
<change beforePath="$PROJECT_DIR$/.idea/workspace.xml" beforeDir="false" afterPath="$PROJECT_DIR$/.idea/workspace.xml" afterDir="false" /> <change beforePath="$PROJECT_DIR$/.idea/workspace.xml" beforeDir="false" afterPath="$PROJECT_DIR$/.idea/workspace.xml" afterDir="false" />
<change beforePath="$PROJECT_DIR$/train.py" beforeDir="false" afterPath="$PROJECT_DIR$/train.py" afterDir="false" />
</list> </list>
<ignored path="$PROJECT_DIR$/out/" /> <ignored path="$PROJECT_DIR$/out/" />
<option name="EXCLUDED_CONVERTED_TO_IGNORED" value="true" /> <option name="EXCLUDED_CONVERTED_TO_IGNORED" value="true" />
@ -48,7 +47,7 @@
<counts> <counts>
<entry key="Dockerfile" value="81" /> <entry key="Dockerfile" value="81" />
<entry key="md" value="104" /> <entry key="md" value="104" />
<entry key="py" value="2552" /> <entry key="py" value="2553" />
<entry key="sh" value="5" /> <entry key="sh" value="5" />
</counts> </counts>
</usages-collector> </usages-collector>
@ -57,7 +56,7 @@
<entry key="Bash" value="5" /> <entry key="Bash" value="5" />
<entry key="Dockerfile" value="81" /> <entry key="Dockerfile" value="81" />
<entry key="Markdown" value="104" /> <entry key="Markdown" value="104" />
<entry key="Python" value="2552" /> <entry key="Python" value="2553" />
</counts> </counts>
</usages-collector> </usages-collector>
</session> </session>
@ -92,8 +91,8 @@
<file pinned="false" current-in-tab="true"> <file pinned="false" current-in-tab="true">
<entry file="file://$PROJECT_DIR$/train.py"> <entry file="file://$PROJECT_DIR$/train.py">
<provider selected="true" editor-type-id="text-editor"> <provider selected="true" editor-type-id="text-editor">
<state relative-caret-position="164"> <state relative-caret-position="167">
<caret line="112" column="20" lean-forward="true" selection-start-line="112" selection-start-column="20" selection-end-line="112" selection-end-column="20" /> <caret line="106" column="40" lean-forward="true" selection-start-line="106" selection-start-column="40" selection-end-line="106" selection-end-column="40" />
<folding> <folding>
<element signature="e#0#28#0" expanded="true" /> <element signature="e#0#28#0" expanded="true" />
</folding> </folding>
@ -224,7 +223,7 @@
<component name="PropertiesComponent"> <component name="PropertiesComponent">
<property name="WebServerToolWindowFactoryState" value="false" /> <property name="WebServerToolWindowFactoryState" value="false" />
<property name="aspect.path.notification.shown" value="true" /> <property name="aspect.path.notification.shown" value="true" />
<property name="com.android.tools.idea.instantapp.provision.ProvisionBeforeRunTaskProvider.myTimeStamp" value="1588384213991" /> <property name="com.android.tools.idea.instantapp.provision.ProvisionBeforeRunTaskProvider.myTimeStamp" value="1588384735776" />
<property name="go.gopath.indexing.explicitly.defined" value="true" /> <property name="go.gopath.indexing.explicitly.defined" value="true" />
<property name="nodejs_interpreter_path.stuck_in_default_project" value="undefined stuck path" /> <property name="nodejs_interpreter_path.stuck_in_default_project" value="undefined stuck path" />
<property name="nodejs_npm_path_reset_for_default_project" value="true" /> <property name="nodejs_npm_path_reset_for_default_project" value="true" />
@ -265,16 +264,15 @@
<option name="presentableId" value="Default" /> <option name="presentableId" value="Default" />
<updated>1588152877746</updated> <updated>1588152877746</updated>
<workItem from="1588152880522" duration="16973000" /> <workItem from="1588152880522" duration="16973000" />
<workItem from="1588319878551" duration="20056000" /> <workItem from="1588319878551" duration="20451000" />
</task> </task>
<servers /> <servers />
</component> </component>
<component name="TimeTrackingManager"> <component name="TimeTrackingManager">
<option name="totallyTimeSpent" value="37029000" /> <option name="totallyTimeSpent" value="37424000" />
</component> </component>
<component name="ToolWindowManager"> <component name="ToolWindowManager">
<frame x="0" y="0" width="1280" height="800" extended-state="0" /> <frame x="0" y="0" width="1280" height="800" extended-state="0" />
<editor active="true" />
<layout> <layout>
<window_info id="Designer" order="0" /> <window_info id="Designer" order="0" />
<window_info id="UI Designer" order="1" /> <window_info id="UI Designer" order="1" />
@ -396,8 +394,8 @@
</entry> </entry>
<entry file="file://$PROJECT_DIR$/train.py"> <entry file="file://$PROJECT_DIR$/train.py">
<provider selected="true" editor-type-id="text-editor"> <provider selected="true" editor-type-id="text-editor">
<state relative-caret-position="164"> <state relative-caret-position="167">
<caret line="112" column="20" lean-forward="true" selection-start-line="112" selection-start-column="20" selection-end-line="112" selection-end-column="20" /> <caret line="106" column="40" lean-forward="true" selection-start-line="106" selection-start-column="40" selection-end-line="106" selection-end-column="40" />
<folding> <folding>
<element signature="e#0#28#0" expanded="true" /> <element signature="e#0#28#0" expanded="true" />
</folding> </folding>

109
train.py
View File

@ -2,7 +2,6 @@ from pandas import DataFrame
from pandas import Series from pandas import Series
from pandas import concat from pandas import concat
from pandas import read_csv from pandas import read_csv
from pandas import datetime
from sklearn.metrics import mean_squared_error from sklearn.metrics import mean_squared_error
from sklearn.preprocessing import MinMaxScaler from sklearn.preprocessing import MinMaxScaler
from keras.models import Sequential from keras.models import Sequential
@ -12,18 +11,13 @@ from math import sqrt
import numpy import numpy
# date-time parsing function for loading the dataset
def parser(x):
return datetime.strptime('190' + x, '%Y-%m')
# frame a sequence as a supervised learning problem # frame a sequence as a supervised learning problem
def timeseries_to_supervised(data, lag=1): def timeseries_to_supervised(data, lag=1):
df = DataFrame(data) df = DataFrame(data)
columns = [df.shift(i) for i in range(1, lag + 1)] columns = [df.shift(i) for i in range(1, lag + 1)]
columns.append(df) columns.append(df)
df = concat(columns, axis=1) df = concat(columns, axis=1)
df.fillna(0, inplace=True) df = df.drop(0)
return df return df
@ -56,8 +50,8 @@ def scale(train, test):
# inverse scaling for a forecasted value # inverse scaling for a forecasted value
def invert_scale(scaler, X, value): def invert_scale(scaler, X, yhat):
new_row = [x for x in X] + [value] new_row = [x for x in X] + [yhat]
array = numpy.array(new_row) array = numpy.array(new_row)
array = array.reshape(1, len(array)) array = array.reshape(1, len(array))
inverted = scaler.inverse_transform(array) inverted = scaler.inverse_transform(array)
@ -73,7 +67,6 @@ def fit_lstm(train, batch_size, nb_epoch, neurons):
model.add(Dense(1)) model.add(Dense(1))
model.compile(loss='mean_squared_error', optimizer='adam') model.compile(loss='mean_squared_error', optimizer='adam')
for i in range(nb_epoch): for i in range(nb_epoch):
print("Epoch {}/{}".format(i, nb_epoch))
model.fit(X, y, epochs=1, batch_size=batch_size, verbose=0, shuffle=False) model.fit(X, y, epochs=1, batch_size=batch_size, verbose=0, shuffle=False)
model.reset_states() model.reset_states()
return model return model
@ -86,44 +79,62 @@ def forecast_lstm(model, batch_size, X):
return yhat[0, 0] return yhat[0, 0]
# run a repeated experiment
def experiment(repeats, series, seed):
# transform data to be stationary
raw_values = series.values
diff_values = difference(raw_values, 1)
# transform data to be supervised learning
supervised = timeseries_to_supervised(diff_values, 1)
supervised_values = supervised.values
# split data into train and test-sets
train, test = supervised_values[0:-12], supervised_values[-12:]
# transform the scale of the data
scaler, train_scaled, test_scaled = scale(train, test)
# run experiment
error_scores = list()
for r in range(repeats):
# fit the model
batch_size = 4
train_trimmed = train_scaled[2:, :]
lstm_model = fit_lstm(train_trimmed, batch_size, 3000, 4)
# forecast the entire training dataset to build up state for forecasting
if seed:
train_reshaped = train_trimmed[:, 0].reshape(len(train_trimmed), 1, 1)
lstm_model.predict(train_reshaped, batch_size=batch_size)
# forecast test dataset
test_reshaped = test_scaled[:, 0:-1]
test_reshaped = test_reshaped.reshape(len(test_reshaped), 1, 1)
output = lstm_model.predict(test_reshaped, batch_size=batch_size)
predictions = list()
for i in range(len(output)):
yhat = output[i, 0]
X = test_scaled[i, 0:-1]
# invert scaling
yhat = invert_scale(scaler, X, yhat)
# invert differencing
yhat = inverse_difference(raw_values, yhat, len(test_scaled) + 1 - i)
# store forecast
predictions.append(yhat)
# report performance
rmse = sqrt(mean_squared_error(raw_values[-12:], predictions))
print('%d) Test RMSE: %.3f' % (r + 1, rmse))
error_scores.append(rmse)
return error_scores
# load dataset # load dataset
series = read_csv('data.csv', header=0, index_col=0, squeeze=True) series = read_csv('data.csv', header=0, index_col=0, squeeze=True)
# experiment
# transform data to be stationary repeats = 30
raw_values = series.values results = DataFrame()
diff_values = difference(raw_values, 1) # with seeding
with_seed = experiment(repeats, series, True)
# transform data to be supervised learning results['with-seed'] = with_seed
supervised = timeseries_to_supervised(diff_values, 1) # without seeding
supervised_values = supervised.values without_seed = experiment(repeats, series, False)
results['without-seed'] = without_seed
# split data into train and test-sets # summarize results
train, test = supervised_values[0:-12], supervised_values[-12:] print(results.describe())
# save boxplot
# transform the scale of the data results.boxplot()
scaler, train_scaled, test_scaled = scale(train, test)
# fit the model
lstm_model = fit_lstm(train_scaled, 1, 30, 4)
# forecast the entire training dataset to build up state for forecasting
train_reshaped = train_scaled[:, 0].reshape(len(train_scaled), 1, 1)
lstm_model.predict(train_reshaped, batch_size=1)
# walk-forward validation on the test data
predictions = list()
for j in range(len(test_scaled)):
# make one-step forecast
X, y = test_scaled[j, 0:-1], test_scaled[j, -1]
yhat = forecast_lstm(lstm_model, 1, X)
# invert scaling
yhat = invert_scale(scaler, X, yhat)
# invert differencing
yhat = inverse_difference(raw_values, yhat, len(test_scaled) + 1 - j)
# store forecast
predictions.append(yhat)
expected = raw_values[len(train) + j + 1]
print('Month=%d, Predicted=%f, Expected=%f' % (j + 1, yhat, expected))
# report performance
rmse = sqrt(mean_squared_error(raw_values[-12:], predictions))
print('Test RMSE: %.3f' % rmse)