---
title: "Aesera"
description: "This page explains how to build, train, deploy and store Aesera models."
canonical: https://www.quantconnect.com/docs/v2/writing-algorithms/machine-learning/popular-libraries/aesera
markdown: https://www.quantconnect.com/docs/v2/writing-algorithms/machine-learning/popular-libraries/aesera.md
language: python
source: https://github.com/QuantConnect/Documentation/tree/master/03%20Writing%20Algorithms/31%20Machine%20Learning/03%20Popular%20Libraries/01%20Aesera
modified: 2026-10-30
index: https://www.quantconnect.com/llms.txt
---

# Aesera

[Documentation](https://www.quantconnect.com/docs/v2.md) > [Writing Algorithms](https://www.quantconnect.com/docs/v2/writing-algorithms.md) > Machine Learning > [Popular Libraries](https://www.quantconnect.com/docs/v2/writing-algorithms/machine-learning/popular-libraries.md) > Aesera

> Code examples are shown in Python. Append `?lang=csharp` to this URL for the C# version of this page.

## Introduction

This page explains how to build, train, deploy and store `Aesera` models.

> To access Aesara, use the Legacy package environment. To set this environment, see the documentation in [Cloud Platform](https://www.quantconnect.com/docs/v2/cloud-platform/projects/package-environments.md#02-Set-Environment) or [Local Platform](https://www.quantconnect.com/docs/v2/local-platform/development-environment/packages-and-libraries.md#02-Set-Environment).

## Import Libraries

Import the `aesera` and `sklearn` libraries.

```python
from AlgorithmImports import *
import aesara
import aesara.tensor as at
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import MinMaxScaler
import joblib
```

You need the `joblib` library to store models.

## Create Subscriptions

In the `initialize` method, [subscribe to some data](https://www.quantconnect.com/docs/v2/writing-algorithms/securities/asset-classes.md) so you can train the `sklearn` model and make predictions.

```python
# Add a security and save a reference to its Symbol.
self._symbol = self.add_equity("SPY", Resolution.DAILY).symbol
```

## Build Models

In this example, build a logistic regression prediction model that uses the following features and labels:

| Data Category | Description |
| --- | --- |
| Features | Normalized daily close price of the SPY over the last 5 days |
| Labels | Return direction of the SPY over the next day |

The following image shows the time difference between the features and labels:

![Features and labels for training](https://cdn.quantconnect.com/i/tu/ml-keras-function.png)

Follow the below steps to build the model:

1. Initialize variables.
   ```python
   # Declare Aesara's symbolic variables.
   x = at.dmatrix("x")
   y = at.dvector("y")

   # Initialize a NumPy random number generator with a fixed seed for reproducibility.
   rng = np.random.default_rng(100)

   # Initialize the weight vector w randomly using shared so the model coefficients keep their values
   # between training iterations (updates).
   w = aesara.shared(rng.standard_normal(5), name="w")

   # Initialize the bias term.
   b = aesara.shared(0., name="b")
   ```
2. Construct the model graph.
   ```python
   # Construct the Aesara expression graph.
   p_1 = 1 / (1 + at.exp(-at.dot(x, w) - b))           # Logistic transformation.
   prediction = p_1 > 0.5                              # The prediction thresholded.
   xent = y * at.log(p_1) - (1 - y) * at.log(1 - p_1)  # Cross-entropy log-loss function.
   cost = xent.mean() + 0.01 * (w ** 2).sum()          # The cost to minimize (MSE).
   gw, gb = at.grad(cost, [w, b])                      # Compute the gradient of the cost.
   ```
3. Compile the model.
   ```python
   # Define training and prediction functions with Aesara to optimize weights
   # and biases during training and generate predictions.
   self._train = aesara.function(
       inputs=[x, y],
       outputs=[prediction, xent],
       # Update weights and biases.
       updates=((w, w - 0.1 * gw), (b, b - 0.1 * gb)))
   self.predict = aesara.function(inputs=[x], outputs=prediction)
   ```

## Train Models

You can train the model at the beginning of your algorithm and you can periodically re-train it as the algorithm executes.

### Warm Up Training Data

You need historical data to initially train the model at the start of your algorithm. To get the initial training data, in the `initialize` method, make a [history request](https://www.quantconnect.com/docs/v2/writing-algorithms/historical-data/history-requests.md).

```python
# Fill a RollingWindow with 2 years of training data.
training_length = 252*2
self.training_data = RollingWindow(training_length)
history = self.history[TradeBar](self._symbol, training_length, Resolution.DAILY)
for trade_bar in history:
    self.training_data.add(trade_bar)
```

### Define a Training Method

To train the model, define a method that fits the model with the training data.

```python
# Prepare feature and label data for training.
def get_features_and_labels(self, n_steps=5):
    training_df = self.pandas_converter.get_data_frame[TradeBar](list(self.training_data)[::-1])['close']

    features = []
    for i in range(1, n_steps + 1):
        close = training_df.shift(i)[n_steps:-1]
        close.name = f"close-{i}"
        features.append(close)
    features = pd.concat(features, axis=1)
    # Normalize using the 5 day interval.
    features = MinMaxScaler().fit_transform(features.T).T[4:]

    Y = training_df.pct_change().shift(-1)[n_steps*2-1:-1].reset_index(drop=True)
    labels = np.array([1 if y > 0 else 0 for y in Y])   # Binary class.

    return features, labels

def my_training_method(self):
    features, labels = self.get_features_and_labels()
    D = (features, labels)
    self._train(D[0], D[1])
```

### Set Training Schedule

To train the model at the beginning of your algorithm, in the `initialize` method, call the `train` method.

```python
# Train the model right now.
self.train(self.my_training_method)
```

To periodically re-train the model as your algorithm executes, in the `initialize` method, call the `train` method as a [Scheduled Event](https://www.quantconnect.com/docs/v2/writing-algorithms/scheduled-events.md).

```python
# Train the model every Sunday at 8:00 AM.
self.train(self.date_rules.every(DayOfWeek.SUNDAY), self.time_rules.at(8, 0), self.my_training_method)
```

### Update Training Data

To update the training data as the algorithm executes, in the `on_data` method, add the current `TradeBar` to the `RollingWindow` that holds the training data.

```python
# Add the current bar to the training data to ensure the model trains with the most recent market data.
def on_data(self, slice: Slice) -> None:
    if self._symbol in slice.bars:
        self.training_data.add(slice.bars[self._symbol])
```

## Predict Labels

To predict the labels of new data, in the `on_data` method, get the most recent set of features and then call the `predict` method.

```python
# Get the feature set and make a prediction.
features, _ = self.get_features_and_labels()
prediction = self.predict(features[-1].reshape(1, -1))
prediction = float(prediction)
```

You can use the label prediction to place orders.

```python
# Place orders based on the prediction.
if prediction == 1:
    self.set_holdings(self._symbol, 1)
elif prediction == 0:
    self.set_holdings(self._symbol, -1)
```

## Save Models

Follow these steps to save `sklearn` models into the [Object Store](https://www.quantconnect.com/docs/v2/writing-algorithms/object-store.md):

1. Set the key name you want to store the model under in the Object Store.
   ```python
   # Set the storage key to something representative.
   model_key = f"{self.project_id}/model"
   ```
2. Call the `get_file_path` method with the key.
   ```python
   # Get the file path to save and access the model in the Object Store.
   file_name = self.object_store.get_file_path(model_key)
   ```
   This method returns the file path where the model will be stored.
3. Call the `dump` method the file path.
   ```python
   # Serialize and save the model.
   joblib.dump(self.predict, file_name)
   ```
   If you dump the model using the `joblib` module before you save the model, you don't need to retrain the model.

## Upload Models

If you train models locally or in another environment, you can upload the model files to the Object Store so your algorithms can use them. Use the model key that matches the key your algorithm expects when it calls `object_store.get_file_path`.

Follow one of these approaches to upload your model files:

### LEAN CLI

Run the [`lean cloud object-store set`](https://www.quantconnect.com/docs/v2/lean-cli/api-reference/lean-cloud-object-store-set.md) command to upload a local file to the Object Store.

```
$ lean cloud object-store set <projectId>/model <pathTo>/model
```

Replace `<projectId>` with your project Id and `<pathTo>` with the path to the local model file.

### Cloud API

Use the [Upload Object Store Files](https://www.quantconnect.com/docs/v2/cloud-platform/api-reference/object-store-management/upload-object-store-files.md) endpoint to upload a model file through the API.

## Load Models

You can load and trade with pre-trained `sklearn` models that you saved in the Object Store. To load a `sklearn` model from the Object Store, in the `initialize` method, get the file path to the saved model and then call the `load` method.

```python
# Load the trained Aesera model from the Object Store.
def initialize(self) -> None:
    model_key = f"{self.project_id}/model"
    if self.object_store.contains_key(model_key):
        file_name = self.object_store.get_file_path(model_key)
        self.model = joblib.load(file_name)
```

The `contains_key` method returns a boolean that represents if the `model_key` is in the Object Store. If the Object Store does not contain the `model_key`, save the model using the `model_key` before you proceed.

## Examples

The following examples demonstrate some common practices for using `Aesera` library.

### Example 1: Aesera

The below algorithm makes use of `Aesera` library to predict the future price movement using the previous 5 OHLCV data. The model is trained using rolling 2-year data. To ensure the model applicable to the current market environment, we recalibrate the model on every Sunday.

```python
import aesara
import aesara.tensor as at
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import MinMaxScaler
import joblib

class AeseraExampleAlgorithm(QCAlgorithm):
    def initialize(self) -> None:
        self.set_start_date(2024, 9, 1)
        self.set_end_date(2024, 12, 31)
        self.set_cash(100000)
        # Request SPY data for model training, prediction and trading.
        self._symbol = self.add_equity("SPY", Resolution.DAILY).symbol

        # Use the session rolling window to cache 2-year daily data for training.
        training_length = 252*2
        self.securities[self._symbol].session.size = training_length
        self.set_warm_up(training_length, Resolution.DAILY)

        # Declare Aesara symbolic variables.
        x = at.dmatrix("x")
        y = at.dvector("y")

        # Initialize the weight vector w randomly using shared so the model coefficients keep
        # their values between training iterations (updates).
        rng = np.random.default_rng(100)
        w = aesara.shared(rng.standard_normal(5), name="w")
        # Initialize the bias term.
        b = aesara.shared(0., name="b")

        # Construct Aesara expression graph.
        p_1 = 1 / (1 + at.exp(-at.dot(x, w) - b))       # Probability that target = 1
        prediction = p_1 > 0.5                          # The prediction thresholded
        xent = y * at.log(p_1) - (1-y) * at.log(1-p_1)  # Cross-entropy log-loss function
        cost = xent.mean() + 0.01 * (w ** 2).sum()      # The cost to minimize
        gw, gb = at.grad(cost, [w, b])                  # Compute the gradient of the cost

        # Compile the model and train it later.
        self._train = aesara.function(
                inputs=[x, y],
                outputs=[prediction, xent],
                updates=((w, w - 0.1 * gw), (b, b - 0.1 * gb)))
        self.predict = aesara.function(inputs=[x], outputs=prediction)

    def on_warmup_finished(self) -> None:
        # Train the model to use the prediction right away.
        self.train(self.my_training_method)
        # Recalibrate the model weekly to ensure its accuracy on the updated domain.
        self.train(self.date_rules.every(DayOfWeek.SUNDAY), self.time_rules.at(8,0), self.my_training_method)

    def get_features_and_labels(self, n_steps=5) -> None:
        # Train and predict the return data, which is more normalized and stationary.
        training_df = self.pandas_converter.get_data_frame[TradeBar](list(self.securities[self._symbol].session)[::-1])['close']

        # Stack the data for 5-day OHLCV data per each sample to train with.
        features = []
        for i in range(1, n_steps + 1):
            close = training_df.shift(i)[n_steps:-1]
            close.name = f"close-{i}"
            features.append(close)
        features = pd.concat(features, axis=1)
        # Normalize using the 5 day interval.
        features = MinMaxScaler().fit_transform(features.T).T[4:]

        Y = training_df.pct_change().shift(-1)[n_steps*2-1:-1].reset_index(drop=True)
        # Train to predict the direction only.
        labels = np.array([1 if y > 0 else 0 for y in Y])   # binary class

        return features, labels

    def my_training_method(self) -> None:
        # Prepare the processed training data.
        features, labels = self.get_features_and_labels()
        D = (features, labels)
        # Recalibrate the model based on updated data.
        self._train(D[0], D[1])

        # Store the model to object store to retrieve it in other instances in case the algorithm stops.
        model_key = f"{self.project_id}/model_test_aesera"
        file_name = self.object_store.get_file_path(model_key)
        joblib.dump(self.predict, file_name)
        self.object_store.save(model_key)

    def on_data(self, slice: Slice) -> None:
        if self.is_warming_up or self._symbol not in slice.bars:
            return

        # Get prediction using the latest 5 OHLCV.
        features, _ = self.get_features_and_labels()
        prediction = self.predict(features[-1].reshape(1, -1))
        prediction = float(prediction)

        # If the predicted direction is going upward, buy SPY.
        if prediction == 1:
            self.set_holdings(self._symbol, 1)
        # If the predicted direction is going downward, sell SPY.
        elif prediction == 0:
            self.set_holdings(self._symbol, -1)
```