Thanks for your support, I really appreciate, together we are better! 🙂

This is how I solved the step four issue:

84d48667-ba93-4ee2-8e09-413d97179593-image.png

Kind Regards
Eddie

@theflyingdutchman Yes, we are integrating new data sources for a new asset class, once we are done (next week) the data and leaderboard updates will start again.

@support

This is a possible fix, but no gurantee. You have to adjust also the prediction function.

def train_model(data):
"""Create and train the models working on an asset-by-asset basis."""

models = dict()

asset_name_all = data.coords['asset'].values

data = data.sel(time=slice('2013-05-01',None)) # cut the noisy data head before 2013-05-01

features_all = get_features(data)
target_all = get_target_classes(data)

model = create_model()

for asset_name in asset_name_all:

# drop missing values: target_cur = target_all.sel(asset=asset_name).dropna('time', 'any') features_cur = features_all.sel(asset=asset_name).dropna('time', 'any') # align features and targets: target_for_learn_df, feature_for_learn_df = xr.align(target_cur, features_cur, join='inner') if len(features_cur.time) < 10: # not enough points for training continue try: model.fit(feature_for_learn_df.values, target_for_learn_df) models[asset_name] = model except KeyboardInterrupt as e: raise e except: logging.exception('model training failed')

return models

@alfredaita Thanks, can it maybe help? https://stackoverflow.com/questions/53421626/valueerror-found-array-with-0-sample-s-shape-0-1-while-a-minimum-of-1-is

@support Thanks for the clarification!

I have a lot of strategies in queue. If the strategies are not finished running by the deadline, will they not be considered?

Hello colleagues.

The solution in case of predicting one financial instrument can be the following (train_period changed)

def load_data(period): return qndata.cryptodaily_load_data(tail=period, assets=['BTC']) def train_model(data): """ train the LSTM network """ asset_name = 'BTC' features_all = get_features(data) target_all = get_target_classes(data) model = get_model() # drop missing values: target_cur = target_all.sel(asset=asset_name).dropna('time', 'any') features_cur = features_all.sel(asset=asset_name).dropna('time', 'any') # align features and targets: target_for_learn_df, feature_for_learn_df = xr.align(target_cur, features_cur, join='inner') criterion = nn.MSELoss() # define loss function optimiser = optim.LBFGS(model.parameters(), lr=0.08) # we use an LBFGS solver as optimiser epochs = 1 # how many epochs for i in range(epochs): def closure(): # reevaluates the model and returns the loss (forward pass) optimiser.zero_grad() # input tensor in_ = torch.zeros(1, len(feature_for_learn_df.values)) in_[0, :] = torch.tensor(np.array(feature_for_learn_df.values)) # output out = model(in_) # target tensor target = torch.zeros(1, len(target_for_learn_df.values)) target[0, :] = torch.tensor(np.array(target_for_learn_df.values)) # evaluate loss loss = criterion(out, target) loss.backward() return loss optimiser.step(closure) # updates weights return model weights = qnbt.backtest_ml( load_data=load_data, train=train_model, predict=predict, train_period=1 * 365, # the data length for training in calendar days retrain_interval=365, # how often we have to retrain models (calendar days) retrain_interval_after_submit=1, # how often retrain models after submission during evaluation (calendar days) predict_each_day=False, # Is it necessary to call prediction for every day during backtesting? # Set it to true if you suspect that get_features is looking forward. competition_type='crypto_daily_long_short', # competition type lookback_period=365, # how many calendar days are needed by the predict function to generate the output start_date='2014-01-01', # backtest start date build_plots=True # do you need the chart? )

@antinomy thanks!

@alfredaita
In case you don't want to run init.py every time in order to install external libraries, I came up with a solution for this. You basically install the library in a folder in your home directory and let the strategy create symlinks to the module path at runtime. More details in this post.

Hi @support,

I modified the retraining interval to 1 day and it worked. Thank you for the help.

You guys rock!

@anshul96go Dear Anshul, it means that the weights for each day have to be generated in less than 10 minutes of time per day.

Note that all submissions are processed on the server after submission using a muti-pass approach (not single-pass).

10 minutes per day, times 250 days, times 10 years, that is more than 400 hours of running time.

@magenta-grimer Hello, you can have at most 50 running submissions in your user area. You can stop any of them any moment and replace it with another one.

Before the end of the Q17 submission phase, you should select at most 15 of them. These will take part to the live contest.

@cyan-graveler Yes, that is correct. Last year we had to replace the futures data for forex with spot data, therefore some fields are missing. More details can be found at:

https://quantiacs.com/community/topic/60/data-for-futures

@support Thanks seems fine

@antinomy Hello, sorry for delay again. We found a problem with the data provider, sorry.

@cespadilla Hello.

The reason is in "train_model" function.

def train_model(data): asset_name_all = data.coords['asset'].values features_all = get_features(data) target_all = get_target_classes(data) models = dict() for asset_name in asset_name_all: # drop missing values: target_cur = target_all.sel(asset=asset_name).dropna('time', 'any') features_cur = features_all.sel(asset=asset_name).dropna('time', 'any') target_for_learn_df, feature_for_learn_df = xr.align(target_cur, features_cur, join='inner') if len(features_cur.time) < 10: continue model = get_model() try: model.fit(feature_for_learn_df.values, target_for_learn_df) models[asset_name] = model except: logging.exception('model training failed') return models

If there are less than 10 features for training the model, then the model is not created (if len(features_cur.time) < 10).

This condition makes sense. I would not remove it.

The second thing that can affect is the retraining interval of the model ("retrain_interval").

weights = qnbt.backtest_ml( train=train_model, predict=predict_weights, train_period=2 *365, # the data length for training in calendar days retrain_interval=10 *365, # how often we have to retrain models (calendar days) retrain_interval_after_submit=1, # how often retrain models after submission during evaluation (calendar days) predict_each_day=False, # Is it necessary to call prediction for every day during backtesting? # Set it to true if you suspect that get_features is looking forward. competition_type='crypto_daily_long_short', # competition type lookback_period=365, # how many calendar days are needed by the predict function to generate the output start_date='2014-01-01', # backtest start date analyze = True, build_plots=True # do you need the chart? )