### Python Suggestion Example

Source: https://systematica.mintlify.app/api-reference/systematica/portfolio/tuners/utils

Illustrative Python code demonstrating how the 'suggest' method might be used within an Optuna optimization loop to get a suggested value for a parameter.

```Python
import optuna

def objective(trial: optuna.trial._trial.Trial):
    # Example: Suggest a learning rate
    learning_rate = trial.suggest_float('learning_rate', 1e-5, 1e-1, log=True)
    
    # Example: Suggest a batch size
    batch_size = trial.suggest_int('batch_size', 16, 128, step=16)
    
    # Example: Suggest an activation function
    activation = trial.suggest_categorical('activation', ['relu', 'tanh', 'sigmoid'])
    
    # ... use these suggested hyperparameters to train a model ...
    accuracy = 0.9 # Placeholder for actual model performance
    return accuracy

# In a real scenario, you would create a study and optimize the objective function
# study = optuna.create_study(direction='maximize')
# study.optimize(objective, n_trials=100)
```

--------------------------------

### OptunaTuner Methods

Source: https://systematica.mintlify.app/api-reference/systematica/portfolio/tuners/optuna_tuner

Documentation for the OptunaTuner class methods, covering objective function creation and Optuna study execution.

```APIDOC
OptunaTuner:
  create_objective(self, feature_config: FeatureConfig) -> Callable
    Creates an optimized objective function for Optuna by separating fixed and tunable parameters once.
    Parameters:
      feature_config (FeatureConfig): Feature configurations.
    Returns:
      tp.Callable: A callable objective function for Optuna optimization.

  run_feature(self, feature_config: FeatureConfig) -> optuna.study.study.Study
    Run feature config.
    Parameters:
      feature_config (FeatureConfig): feature configuration.
    Returns:
      optuna.Study: Optuna study.

  run(self, objective: Callable) -> optuna.study.study.Study
    Run optuna study or tracker.
    Parameters:
      objective (tp.Callable): Objective function.
    Returns:
      study : optuna.Study: Optuna Study.

  run_tracker(self, objective: Callable) -> optuna.study.study.Study
    Run tracker.
    Parameters:
      objective (tp.Callable): Objective function.
    Returns:
      study : optuna.Study: Optuna Study.

  run_optuna(self, objective: Callable) -> optuna.study.study.Study
    Run optuna.
    Parameters:
      objective (tp.Callable): Objective function.
    Returns:
      study : optuna.Study: Optuna Study.
```

--------------------------------

### Sandbox Class Constructor and Parameters

Source: https://systematica.mintlify.app/api-reference/systematica/portfolio/sandbox/sandbox

Initializes the Sandbox class, which manages feature calculation and strategy optimization. It accepts various configurations for Optuna samplers, pruners, metrics, and parallel job execution. Instance variables detail the configurable aspects of the Sandbox.

```APIDOC
Sandbox:
  __init__(feature_registry: List[~FeatureConfig], pruner: optuna.pruners._base.BasePruner = <optuna.pruners._successive_halving.SuccessiveHalvingPruner object>, sampler: optuna.samplers._base.BaseSampler = <optuna.samplers._tpe.sampler.TPESampler object>, direction: str | List[str] = 'maximize', create_study_kwargs: Dict[str, Any] = _Nothing.NOTHING, n_trials: int = 100, n_completed_trials: int = None, n_jobs: int = -1, gc_after_trial: bool = False, optimize_kwargs: Dict[str, Any] = _Nothing.NOTHING, verbose: bool = True, tracker: systematica.portfolio.trackers.base.BaseTracker = None, metrics: str | List[str] = 'sharpe_ratio', metric_registry_kwargs: Dict[str, Any] = _Nothing.NOTHING, cross_validate: bool = False, reduce_func: Callable = CPUDispatcher(<function nanmedian_nb>), use_rolling: bool = False)
    Handles feature calculation and strategy optimization using a registry of feature configurations.

    Parameters:
      feature_registry: List of feature configurations.
      pruner: Optuna pruner. Defaults to SuccessiveHalvingPruner.
      sampler: Optuna sampler. Defaults to TPESampler.
      direction: Optimization direction for Optuna ('maximize' or 'minimize'). Can be a list.
      create_study_kwargs: Additional Optuna study parameters.
      n_trials: Number of optimization trials for Optuna (default: 100).
      n_completed_trials: Number of trials to complete before stopping (default: None).
      n_jobs: Parallel jobs for Optuna (default: -1).
      gc_after_trial: Flag to run garbage collection after each trial (default: False).
      optimize_kwargs: Additional arguments for Optuna optimization.
      verbose: Enable verbose output (default: True).
      tracker: Log and monitor Optuna hyperparameters using a tracker integration.
      metrics: Performance metrics to compute (default: 'sharpe_ratio'). Can be a list.
      metric_registry_kwargs: Additional arguments for metric registry (e.g., window for rolling metrics).
      cross_validate: Whether to use cross-validation analytics (default: False).
      reduce_func: Reduce function when using cross_validate. Defaults to nanmedian_nb.
      use_rolling: Flag to use rolling method via BaseRollingReport (default: False).
```

--------------------------------

### Get Inputs (get_inputs)

Source: https://systematica.mintlify.app/api-reference/systematica/models/meta_model/base

Retrieves the state representation, specifically the input (X) variables, for model training. It supports downsampling for performance optimization and can return results as a NumPy array or Pandas DataFrame. An example demonstrates how to add features to a vbt.Data object.

```APIDOC
get_inputs(self, downsample: str = '1d', to_numpy: bool = False, state_registry: List[systematica.models.meta_model.state_registries.StateConfig] = None) -> pandas.core.frame.DataFrame | numpy.ndarray

Parameters:
  - downsample (str, optional): Resample data before state computation to speed up the process. If None, no resampling is performed. Defaults to '1d' (daily).
  - to_numpy (bool, optional): Output Numpy array if True. Pandas DataFrame object otherwise. The default is False.
  - state_registry (List[systematica.models.meta_model.state_registries.StateConfig], optional): State registry to use. If None, defaults to STATE_REGISTRY. The default is None.

Returns:
  pd.DataFrame | tp.Array2d: Input (exogenous) variables.

Example:
```python
data = sma.load_clean_data("1d")
rets = sma.get_returns(data.close)
data = data.add_feature("rets", rets, missing_index="drop")
```
```

--------------------------------

### OptunaTuner Constructor

Source: https://systematica.mintlify.app/api-reference/systematica/portfolio/tuners/optuna_tuner

Initializes the OptunaTuner class, configuring parameters for hyperparameter optimization using Optuna. It accepts various settings for samplers, pruners, optimization direction, trial counts, parallel jobs, and metrics.

```APIDOC
OptunaTuner:
  __init__(
    pruner: optuna.pruners._base.BasePruner = <optuna.pruners._successive_halving.SuccessiveHalvingPruner object>,
    sampler: optuna.samplers._base.BaseSampler = <optuna.samplers._tpe.sampler.TPESampler object>,
    direction: str | List[str] = 'maximize',
    create_study_kwargs: Dict[str, Any] = _Nothing.NOTHING,
    n_trials: int = 100,
    n_completed_trials: int | None = None,
    n_jobs: int = -1,
    gc_after_trial: bool = False,
    optimize_kwargs: Dict[str, Any] = _Nothing.NOTHING,
    verbose: bool = True,
    tracker: systematica.portfolio.trackers.base.BaseTracker | None = None,
    metrics: str | List[str] = 'sharpe_ratio',
    metric_registry_kwargs: Dict[str, Any] = _Nothing.NOTHING,
    cross_validate: bool = False,
    reduce_func: Callable = CPUDispatcher(<function nanmedian_nb>),
    use_rolling: bool = False,
  )
    Parameters:
      pruner: The pruner strategy for Optuna.
      sampler: The sampler strategy for Optuna.
      direction: The optimization direction ('maximize' or 'minimize'). Can be a list for multiple directions.
      create_study_kwargs: Additional keyword arguments for creating an Optuna study.
      n_trials: The total number of trials to run.
      n_completed_trials: Number of trials to complete before stopping. If None, n_trials is used.
      n_jobs: Number of parallel jobs for Optuna. -1 means using all available CPU cores.
      gc_after_trial: If True, runs garbage collection after each trial to manage memory.
      optimize_kwargs: Additional keyword arguments for the Optuna optimization process.
      verbose: If True, enables verbose output during optimization.
      tracker: An optional tracker for logging Optuna hyperparameter progress.
      metrics: The performance metric(s) to optimize. Defaults to 'sharpe_ratio'.
      metric_registry_kwargs: Additional arguments for the metric registry, e.g., window size.
      cross_validate: If True, uses cross-validation for optimization.
      reduce_func: Function to reduce results when cross_validate is True.
      use_rolling: Flag to use rolling window calculations for metrics.

    Instance Variables:
      target_names: List[str] - Normalized metric names.
      create_study_kwargs: Dict[str, Any] - Additional optuna study parameters.
      cross_validate: bool - Whether to use cross-validation analytics.
      direction: str | List[str] - Optimization direction for Optuna.
      gc_after_trial: bool - Flag to determine whether to run garbage collection after each trial.
      metric_registry_kwargs: Dict[str, Any] - Additional arguments for metric registry.
      metrics: str | List[str] - Performance metrics to compute.
      n_completed_trials: int - Number of trials to complete before stopping.
      n_jobs: int - Parallel jobs for Optuna.
      n_trials: int - Number of optimization trials for Optuna.
      optimize_kwargs: Dict[str, Any] - Additional arguments for Optuna optimization.
      pruner: optuna.pruners._base.BasePruner - Pruner strategy.
      reduce_func: Callable - Reduce function when using cross_validate.
      sampler: optuna.samplers._base.BaseSampler - Sampler strategy.
      tracker: systematica.portfolio.trackers.base.BaseTracker - Tracker for logging Optuna hyperparameter.
      use_rolling: bool - Flag to use rolling method via BaseRollingReport.
      verbose: bool - Enable verbose output.
```

--------------------------------

### Get TradingView Data

Source: https://systematica.mintlify.app/api-reference/systematica/data/loader

Extracts stock data from TradingView based on market capitalization filter. It uses `vbt.TVData.list_symbols` and `vbt.TVData.pull` to retrieve data for symbols matching an exchange pattern, filtering by market cap (>= 10 billion) and start date. Returns a `vbt.TVData` object.

```python
get_tradingview_data(
    exchange_pattern: str,
    **pull_kwargs,
) -> vectorbtpro.data.custom.tv.TVData

Extract stock data from TradingView based on market capitalization filter. This function uses the `vbt.TVData.list_symbols` and `vbt.TVData.pull` methods to retrieve market data from TradingView for symbols matching a specified exchange pattern. The data is filtered by market capitalization (e.i. greater than or equal to 10 billion), and the symbol names are cleaned up. The data is also filtered by the provided start date.

Parameters:

- `exchange_pattern` (str): A pattern for the exchange symbols (e.g., “NASDAQ”).
- `pull_kwargs` (tp.Kwargs): Additional arguments passed to `vbt.TVData.pull` to customize the data fetching.

Returns:

- `vbt.TVData`: A `vbt.TVData` object containing the filtered stock data from TradingView.
```

--------------------------------

### Cross-Validation Loop Example

Source: https://systematica.mintlify.app/api-reference/systematica/models/meta_model/clf_model

This example demonstrates a typical cross-validation loop, fitting a scikit-learn pipeline on training data slices and predicting on test data slices obtained from the splitter.

```python
>>> X_slices, y_slices = splitter.take(X), splitter.take(y)
>>> y_tests = []
>>> y_preds = []
>>> for split in X_slices.index.unique(level="split"):
...     X_train_slice = X_slices[(split, "train")]
...     y_train_slice = y_slices[(split, "train")]
...     X_test_slice = X_slices[(split, "test")]
...     y_test_slice = y_slices[(split, "test")]
...     try:
...         fitted_pipe = pipeline.fit(X_train_slice, y_train_slice)
...         test_pred = fitted_pipe.predict(X_test_slice)
...         test_pred_ser = pd.Series(test_pred, index=y_test_slice.index)
```

--------------------------------

### Get Symbol Parameters

Source: https://systematica.mintlify.app/api-reference/systematica/portfolio/composers/models

Retrieves symbol parameters for the current context. Returns a dictionary containing symbol parameters.

```APIDOC
get_symbol_params:
  signature: get_symbol_params(self,) -> Dict[str, Any]
  description: Get symbol params.
  returns:
    - type: tp.Kwargs
      description: Symbol params.
```

--------------------------------

### run Method - Systematica PortfolioAnalyzer

Source: https://systematica.mintlify.app/api-reference/systematica/portfolio/composers/models

Executes a pipeline with the given data and symbols. This method is part of the Systematica PortfolioAnalyzer and requires input data, two symbols, and an optional flag to use closing prices.

```APIDOC
run Method:
  Signature:
    run(
        self,
        data: vectorbtpro.data.base.Data,
        s1: str,
        s2: str,
        use_close: bool = True,
    ) -> systematica.portfolio.analyzers.base.PortfolioAnalyzer
  Description:
    Execute the pipeline with the given data and symbols.
  Parameters:
    data (vbt.Data): Input data for the pipeline.
    s1 (str): First symbol.
    s2 (str): Second symbol.
    use_close (bool, optional): Whether to use ‘Close’ prices. Defaults to True.
  Returns:
    PortfolioAnalyzer: Portfolio analyzer object with the pipeline results.
```

--------------------------------

### Instance Variables

Source: https://systematica.mintlify.app/api-reference/systematica/api/models/volume_profile

Describes the configurable instance variables for volatility models, covering data partitioning, scoring, and window settings.

```APIDOC
Instance Variables:

- custom_splitter: str | None
  - Description: Custom splitter to use for data partitioning.
  - Defaults to: None
  - Notes: If set, it should be a string that matches a custom splitter function.

- custom_splitter_kwargs: dict | None
  - Description: Additional keyword arguments for the custom splitter.
  - Defaults to: None
  - Notes: If `custom_splitter` is set, this should contain any necessary parameters for the custom splitter function.

- score_scale: float
  - Description: Score scale coefficient.
  - Defaults to: 2.0
  - Notes: This factor scales the final score output.

- short_term_window: int
  - Description: Short term window.
  - Defaults to: 30
  - Notes: This is the window size used for short-term volume calculations.

- splitter: str
  - Description: Default splitter to be used if `custom_splitter` is not passed.
  - Choices: `from_rolling`, `from_expanding`, `from_custom_rolling`, `from_custom_expanding`
  - Defaults to: "from_custom_rolling"

- surge_scale: float
  - Description: Volume surge scale coefficient.
  - Defaults to: 5.0
  - Notes: This factor amplifies the volume surge detection.

- testing_window: int
  - Description: The size of the testing window.
  - Defaults to: 60
  - Notes: This is the window size used for testing the model.

- training_window: int
  - Description: The size of the training window.
  - Defaults to: 365
  - Notes: This is the window size used for training the model.
```

--------------------------------

### Get Plants API

Source: https://systematica.mintlify.app/api-reference/post-plants

Retrieves a list of plants from the store. This endpoint is available for querying plant data.

```APIDOC
GET /plants

Retrieves a list of plants from the store.
```

--------------------------------

### Systematica Portfolio Tuners API Documentation

Source: https://systematica.mintlify.app/api-reference/systematica/portfolio/tuners/optuna_tuner

Documentation for the Tuners module within the Systematica Portfolio, specifically detailing the Optuna Tuner. This section outlines the components and functionalities related to hyperparameter tuning using Optuna.

```APIDOC
Module: systematica.portfolio.tuners.optuna_tuner

Optuna Tuner:
  Description: Integrates Optuna for hyperparameter optimization within the Systematica portfolio.
  Usage: Used for tuning model parameters to improve performance.
  Dependencies: Optuna library.
  Related Components:
    - Tuners (Base)
    - Tuners (Utils)
```

--------------------------------

### Get Plants API

Source: https://systematica.mintlify.app/api-reference/get-plants

Retrieves all plants accessible to the user from the system. Supports filtering by a limit parameter.

```APIDOC
GET /plants

Returns all plants from the system that the user has access to.

Parameters:
  - Authorization (string, header, required): Bearer authentication header of the form `Bearer <token>`, where `<token>` is your auth token.
  - limit (integer, query): The maximum number of results to return.

Responses:
  - 200 (application/json): Plant response. The response is of type `object[]`.
    Example:
    [
      {
        "name": "<string>",
        "tag": "<string>"
      }
    ]
  - 400: Bad Request
```

```cURL
curl --request GET \
  --url http://sandbox.mintlify.com/plants \
  --header 'Authorization: Bearer <token>'
```

--------------------------------

### Get Annualization Factor

Source: https://systematica.mintlify.app/api-reference/systematica/utils/custom_datetime

Calculates the annualization factor based on a given frequency. This is useful for converting periodic data to an annual basis.

```Python
import pandas as pd

def get_ann_factor(freq: pd.Timedelta) -> float:
    """
    Get annualization factor
    
    Parameters:

    freq (pd.Timedelta): Frequency.
    
    Returns:

    float: Annualization factor.
    """
    # Implementation details would go here
    pass

# Example Usage:
# from datetime import timedelta
# daily_freq = timedelta(days=1)
# ann_factor = get_ann_factor(daily_freq)
# print(ann_factor)
```

--------------------------------

### Initialize Blockforce Theme

Source: https://systematica.mintlify.app/api-reference/systematica/utils/blockforce_theme

Initializes and registers the Blockforce Capital theme with Plotly and vectorbtpro. This function sets up the theme with the Blockforce color palette, layout defaults, and trace styles for consistent visualizations.

```python
init_theme()
```

--------------------------------

### Load Binance Data (Raw & Clean)

Source: https://systematica.mintlify.app/api-reference/systematica/data/loader

Enables loading of historical Binance data, supporting both raw and cleaned datasets. Users can specify the timeframe, file suffix, and date range for data retrieval.

```APIDOC
load_binance_data(timeframe: str = '1d', suffix: str = 'h5', start: str = '2018-01-01 UTC', end: str = None) -> vectorbtpro.data.base.Data
  Loads raw Binance data from files.
  Parameters:
    timeframe: Timeframe of the file to be loaded.
    suffix: Extension (suffix) of the file to load (e.g., 'h5').
    start: The start date for the data in UTC string format (e.g., 'YYYY-MM UTC').
    end: The end date for the data in UTC string format.
  Returns:
    A vbt.Data object containing the fetched price data.
```

```APIDOC
load_clean_data(timeframe: str = '1d', suffix: str = 'h5', start: str = '2018-01-01 UTC', end: str = None) -> vectorbtpro.data.base.Data
  Loads cleaned Binance data from files.
  Parameters:
    timeframe: Timeframe of the file to be loaded.
    suffix: Extension (suffix) of the file to load (e.g., 'h5').
    start: The start date for the data in UTC string format (e.g., 'YYYY-MM UTC').
    end: The end date for the data in UTC string format.
  Returns:
    A vbt.Data object containing the fetched price data.
```

--------------------------------

### Metrics Registry: Get All Metrics

Source: https://systematica.mintlify.app/api-reference/systematica/portfolio/metrics/base

Retrieves metric keys from the registry. Supports filtering by tunable status and specific function types (rolling, pf).

```APIDOC
get_all_metrics(self, include_non_tunable: bool = False) -> List[str]
  Get a list of metric keys from the registry.
  Optionally include non-tunable metrics based on user choice.
  Parameters:
    include_non_tunable (bool, default False): Whether to include non-tunable metrics.
  Returns:
    List[str]: List of metric keys.
```

```APIDOC
get_all_rolling_metrics(self, include_non_tunable: bool = False) -> List[str]
  Get a list of metric keys from the registry that have a valid rolling_func.
  Optionally include non-tunable metrics based on user choice.
  Parameters:
    include_non_tunable (bool, default False): Whether to include non-tunable metrics.
  Returns:
    List[str]: List of metric keys with valid (non-NaNs) rolling_func values.
```

```APIDOC
get_all_pf_metrics(self, include_non_tunable: bool = False) -> List[str]
  Get a list of metric keys from the registry that have a valid pf_func.
  Optionally include non-tunable metrics based on user choice.
  Parameters:
    include_non_tunable (bool, default False): Whether to include non-tunable metrics.
  Returns:
    List[str]: List of metric keys with valid (non-NaNs) pf_func values.
```

```APIDOC
get_all_non_tunable_metrics(self) -> List[str]
  Get a list of non-tunable metric keys from the registry.
  Returns:
    List[str]: List of non-tunable metrics.
```

--------------------------------

### Get Scikit-Learn Pipeline Object

Source: https://systematica.mintlify.app/api-reference/systematica/models/meta_model/clf_model

Shows how to create a scikit-learn Pipeline that chains a StandardScaler for preprocessing and a RandomForestClassifier for classification.

```python
>>> pipeline = make_pipeline(
...     StandardScaler(),
...     RandomForestClassifier()
... )
```

--------------------------------

### Optimize Portfolio with run_from_optimize_func

Source: https://systematica.mintlify.app/api-reference/systematica/portfolio/simulation

Sets up and returns a portfolio optimizer instance, enabling custom sampling and optimization functions. It takes historical market data, returns, signals, and a custom optimization function as input, returning a configured PortfolioOptimizer instance ready for backtesting.

```APIDOC
run_from_optimize_func:
  Signature:
    run_from_optimize_func(
        data: vectorbtpro.data.base.Data,
        returns: numpy.ndarray,
        signals: systematica.signals.base.Signals,
        optimize_func: Callable,
        config: Dict[str, Any] = None,
    ) -> vectorbtpro.portfolio.pfopt.base.PortfolioOptimizer
  Description:
    Sets up and returns a portfolio optimizer instance, using custom sampling and optimization functions.
  Parameters:
    data (vectorbtpro.data.base.Data): Historical market data.
    returns (numpy.ndarray): Array of returns to be used for optimization.
    signals (systematica.signals.base.Signals): Generated trading signals.
    optimize_func (Callable): Custom optimization function that takes the returns and signals as input. This function should return a dictionary of optimized parameters or metrics.
    config (Dict[str, Any], optional): Dictionary with optimization parameters. Defaults to None.
  Returns:
    vbt.PortfolioOptimizer: Configured portfolio optimizer instance ready for backtesting.
  Examples:
    >>> strategy = ...
    >>> portfolio_optimizer = strategy.optimizer(
    ...     ...,
    ...     config={'obj': 'Sharpe', 'every': 'M'}
    ... )
    >>> portfolio = portfolio_optimizer.simulate()
```

--------------------------------

### Get Current UTC Time as String

Source: https://systematica.mintlify.app/api-reference/systematica/utils/custom_datetime

Retrieves the current time in UTC and formats it as a string. The string format is ISO 8601 compliant.

```python
current_utc_str() -> str

# Description:
Current time in UTC in string format.
# Returns:
#   str: Current time in UTC as a string in the format %Y-%m-%dT%H:%M:%S.
```

--------------------------------

### Systematica Library Navigation

Source: https://systematica.mintlify.app/api-reference/systematica/models/ou_process

Provides a general overview of the Systematica library's API reference structure, listing key modules and sections. This helps users navigate to different parts of the documentation.

```APIDOC
API Reference Navigation:
- Configuration
- API
- Data
- Models
  - Arbitrage Index
  - Meta Model
  - Momentum
  - Ou Process
  - Range Breakout
  - Volatility
  - Volume Profile
- Portfolio
- Preprocessing
- Signals
- Utils
```

--------------------------------

### ArbitrageClipIndex Model API Documentation

Source: https://systematica.mintlify.app/api-reference/systematica/api/models/arbitrage_index

Detailed API documentation for the ArbitrageClipIndex model, including its constructor signature, parameter descriptions, and static methods like validation.

```APIDOC
systematica.api.models.arbitrage_index

## `ArbitrageClipIndex`

```
ArbitrageClipIndex(
    training_window: int = 365,
    testing_window: int = 60,
    splitter: str = 'from_custom_rolling',
    custom_splitter: str | None = None,
    custom_splitter_kwargs: Dict[str, Any] | None = None,
    copula: str | type = 'auto',
    list_copulas: list[systematica.models.arbitrage_index.base.BaseCopula] | None = None,
    method: str = 'aic',
    lower: float = -1.0,
    upper: float = 1.0,
    center: float = 0.5,
    sequential: bool = False,
)

```

Arbitrage clip index: Marginal probability model. This model computes the accumulated sum of log returns and applies a clipping transformation to the accumulated index. It is useful for analyzing the performance of two assets over time, particularly in the context of arbitrage opportunities. The clipping transformation ensures that the accumulated index remains within specified bounds, preventing extreme values that could skew the analysis. Method generated by attrs for class ArbitrageClipIndex.

### Ancestors

* `systematica.api.base.BaseStatArb`
* `abc.ABC`

### Static methods

#### `is_valid`

```
is_valid(
    model_params: Dict[str, Any],
    signal_params: Dict[str, Any],
)

```

Validate model parameters.

**Parameters**:

| Name | Type | Default | Description |
| --- | --- | --- | --- |
| `model_params` | `tp.Kwargs` | `--` | Model parameters. |
| `signal_params` | `tp.Kwargs` | `--` | Signal parameters. |

**Raises**:

| Type | Description |
| --- | --- |
| `ValueError` | If `long_entries` or `short_entries` are not within the specified bounds. |
```

--------------------------------

### PortfolioComposer Methods

Source: https://systematica.mintlify.app/api-reference/systematica/portfolio/composers/models

PortfolioComposer is used to compose portfolios from trials. It includes methods to retrieve data using a specified loader and to get the parameters used for composition.

```APIDOC
PortfolioComposer:
  Constructor:
    PortfolioComposer(
        feature_config: Dict[str, Any],
        model_params: Dict[str, Any],
    )
  Description:
    Compose portfolio(s) from trials.
  Class variables:
    feature_config (Dict[str, Any]): Configuration for a feature calculation.
    model_params (Dict[str, Any]): A dictionary of optional keyword arguments needed to compose the portfolio.
  Instance variables:
    all_params (Dict[str, Any]): All parameters.

  get_data Method:
    Signature:
      get_data(
          self,
          loader: vectorbtpro.data.base.Data | Callable = None,
          loader_kwargs: Dict[str, Any] = None,
      ) -> vectorbtpro.data.base.Data
    Description:
      Get data.
    Parameters:
      loader (vbt.Data | Callable, optional): Data loader function or object. If None, uses `load_clean_data`.
      loader_kwargs (tp.Kwargs, optional): Additional keyword arguments for the loader function. If None, uses `config.LOADER_PARAMS`.
    Returns:
      tp.Kwargs: Data params.

  get_compose_params Method:
    Signature:
      get_compose_params(
          self,
      ) -> Dict[str, Any]
    Description:
      Get compose parameters.
    Returns:
      tp.Kwargs: Compose params.
```

--------------------------------

### Systematica API Reference Navigation

Source: https://systematica.mintlify.app/api-reference/systematica/models/volatility

Provides navigation links to various sections of the Systematica API reference, including configuration, data, models, portfolio, preprocessing, signals, and utilities.

```APIDOC
/api-reference/systematica

- Configuration
- API
- Data
- Models
- Portfolio
- Preprocessing
- Signals
- Utils
```

--------------------------------

### Get Current UTC Time

Source: https://systematica.mintlify.app/api-reference/systematica/utils/custom_datetime

Retrieves the current time in Coordinated Universal Time (UTC). Returns a datetime object representing the current UTC timestamp.

```python
current_utc() -> datetime.datetime

# Description:
Current time in UTC.
# Returns:
#   datetime: Current time in UTC as a datetime object.
```

--------------------------------

### run_pipeline Method

Source: https://systematica.mintlify.app/api-reference/systematica/api/models/meta_model

Executes the portfolio pipeline for streamlined backtesting. This method is highly parametrizable, allowing users to configure various aspects of the simulation, from symbol selection and metric computation to order execution details and risk management parameters.

```APIDOC
run_pipeline(
    cls,
    select_symbols: str | int | list[str | int] = None,
    metrics: str | List[str] = None,
    use_rolling: bool = False,
    metric_registry_kwargs: Dict[str, Any] = None,
    size: float | numpy.ndarray = 50,
    size_type: str | numpy.ndarray = 'valuepercent100',
    allow_partial: float | numpy.ndarray = None,
    fees: float | numpy.ndarray = None,
    fixed_fees: float | numpy.ndarray = None,
    slippage: float | numpy.ndarray = None,
    min_size: float | numpy.ndarray = None,
    max_size: float | numpy.ndarray = None,
    size_granularity: float | numpy.ndarray = None,
    sl_stop: float | numpy.ndarray = None,
    tsl_stop: float | numpy.ndarray = None,
    tp_stop: float | numpy.ndarray = None,
    accumulate: bool | str | numpy.ndarray = None,
    init_cash: float | numpy.ndarray = None,
    cash_sharing: bool = True,
    group_by: bool = True,
    call_seq: str | bool = 'auto',
    fill_pos_info: bool = False,
    log: bool | numpy.ndarray = None,
    portfolio_kwargs: Dict[str, Any] = None,
    to_numpy: bool = False,
    validate_metrics: bool = True,
    **model_params
) -> vectorbtpro.portfolio.base.Portfolio | collections.OrderedDict | pandas.core.frame.DataFrame | numpy.ndarray

Description:
  Executes the portfolio pipeline. Method is parametrizable. See `BaseStatArb.run_pipeline`
  Combines scoring, signal generation, and portfolio simulation into a single pipeline for streamlined backtesting.

Parameters:
  - cls: The class instance.
  - select_symbols: Symbol name (e.g., `BTCUSDT` or `ETHUSDT`) or column index (e.g., `0` or `1`) to keep. If `None`, no filtering.
  - metrics: Performance metrics to compute. Defaults to `sharpe_ratio`.
  - use_rolling: Flag to use rolling method via `BaseRollingReport`. Note that both rolling and cv models can be passed. Fallbacks to `BaseReportCV` if model is not using `from_rolling` splitter without error message. Defaults to `False`.
  - metric_registry_kwargs: Parameters for metric registry. If `None`, uses default parameters based on the model type. This is used to specify parameters such as `window` or `minp` for rolling metrics. If using cross-validation, this should be set to `None` as it is not applicable. If using rolling metrics, this should be set to a dictionary with parameters for the metric registry. For example, `{"window": 20}`.
  - size: Size of each order. Defaults to `50`.
  - size_type: Type of size. Defaults to `valuepercent100`.
  - allow_partial: Whether to allow partial fills. Defaults to `None`.
  - fees: Fees per order. Defaults to `None`.
  - fixed_fees: Fixed fees per order. Defaults to `None`.
  - slippage: Slippage per order. Defaults to `None`.
  - min_size: Minimum size of each order. Defaults to `None`.
  - max_size: Maximum size of each order. Defaults to `None`.
  - size_granularity: Granularity of size. Defaults to `None`.
  - sl_stop: Stop-loss level. Defaults to `None`.
  - tsl_stop: Trailing stop-loss level. Defaults to `None`.
  - tp_stop: Take-profit level. Defaults to `None`.
  - accumulate: Whether to accumulate orders. Defaults to `None`.
  - init_cash: Initial cash. Defaults to `None`.
  - cash_sharing: Whether to share cash between assets. Defaults to `True`.
  - group_by: Whether to group by symbol. Defaults to `True`.
  - call_seq: Call sequence. Defaults to `auto`.
  - fill_pos_info: Whether to fill position info. Defaults to `False`.
  - log: Whether to log orders. Defaults to `None`.
  - portfolio_kwargs: Additional keyword arguments for the portfolio simulation.
  - to_numpy: If `True`, returns a `np.ndarray`. Default is `False`, returning a `pd.DataFrame`.
  - validate_metrics: Check if `metric` is valid in the registry. Defaults to `True`.
  - model_params: Additional parameters for `model_func`.

Returns:
  - Backtested portfolio or requested metrics. Types can be `vbt.PF`, `OrderedDict`, `pd.DataFrame`, or `tp.Array`.
```

--------------------------------

### Pie Class Example Usage

Source: https://systematica.mintlify.app/api-reference/systematica/utils/custom_plots

Demonstrates how to instantiate the Pie class with sample data and trace names, and then display the generated figure. Requires numpy and systematica libraries.

```python
>>> import numpy as np
>>> import systematica as sma
>>> pie = sma.Pie(
...     data=np.array([0.06, 0.02, 0.05, 0.09]),
...     trace_names=['SPY', 'TLT', 'XLF', 'XLE'],
... )
>>> pie.fig.show()
```

--------------------------------

### Systematica run() Method

Source: https://systematica.mintlify.app/api-reference/systematica/portfolio/trial_selectors/models

Executes the efficient frontier projection selection algorithm. This method is common across various portfolio optimization strategies within the Systematica library. It requires a pandas DataFrame as input and returns a pandas Series representing the optimal parameter combination.

```APIDOC
run(self, df: pandas.core.frame.DataFrame) -> pandas.core.series.Series
  Description: Execute the efficient frontier projection selection algorithm.
  Parameters:
    - df (pd.DataFrame): DataFrame containing parameters to select from.
  Returns:
    - pd.Series: The parameter combination with maximum Sharpe ratio.
  Dependencies:
    - pandas
  Usage:
    This method is typically called on an instance of a specific optimization strategy class (e.g., ElbowPoint, ClosestIdealPoint, etc.) after initialization.
```

--------------------------------

### Get Classification Report

Source: https://systematica.mintlify.app/api-reference/systematica/models/meta_model/base

Generates a comprehensive classification report that includes precision, recall, F1-score, and support for each class. This is useful for evaluating the performance of a classification model across different classes.

```APIDOC
get_report(self, y_pred: pandas.core.series.Series) -> pandas.core.frame.DataFrame

  Description:
    Build a report showing the main classification metrics.

  Parameters:
    - y_pred (pd.Series): Estimated targets as returned by a classifier.

  Returns:
    pd.DataFrame: DataFrame summary of the precision, recall, F1 score for each class.

  See Also:
    - precision_recall_fscore_support
    - confusion_matrix
    - multilabel_confusion_matrix
```

--------------------------------

### Anthropic Integration Example

Source: https://systematica.mintlify.app/api-reference/systematica/portfolio/analyzers

Indicates potential integration with Anthropic's Claude AI model within the Systematica framework, likely for natural language processing or analysis tasks.

```Python
# Example of how Anthropic's Claude might be used within Systematica
# This is a conceptual representation based on the text.

# Assuming a class structure like:
# class AnthropicAnalyzer(systematica.portfolio.analyzers.base.BaseAnalyzer):
#     def __init__(self, model_name: str = 'claude-3-opus-20240229'):
#         super().__init__()
#         self.model_name = model_name
#         # Initialize Anthropic client here
#         # self.client = AnthropicClient(api_key='YOUR_API_KEY')

#     def analyze(self, data: dict) -> dict:
#         # Prepare prompt for Claude
#         prompt = f"Analyze the following portfolio data: {data}"
#         # Call Anthropic API
#         # response = self.client.messages.create(
#         #     model=self.model_name,
#         #     max_tokens=1000,
#         #     messages=[
#         #         {"role": "user", "content": prompt}
#         #     ]
#         # )
#         # Process Claude's response
#         # return {'analysis': response.content[0].text}
#         pass

# Example Usage (conceptual):
# analyzer = AnthropicAnalyzer()
# portfolio_data = {'returns': [0.01, 0.02, -0.005], 'volatility': 0.05}
# result = analyzer.analyze(portfolio_data)
# print(result)

# The text 'AnthropicOpen in Claude' suggests this capability exists or is being documented.
```

--------------------------------

### Run Study Method - Python API

Source: https://systematica.mintlify.app/api-reference/systematica/portfolio/trackers/base

Executes an optimization study using a provided objective function and configuration. It takes keyword arguments for study creation and optimization, returning the completed Optuna study object.

```APIDOC
run_study:
  signature: run_study(self, objective: Callable, create_study_kwargs: Dict[str, Any], optimize_kwargs: Dict[str, Any]) -> optuna.study.study.Study
  description: Run study.
  parameters:
    - name: objective
      type: tp.Callable
      description: The objective function to be optimized.
    - name: create_study_kwargs
      type: tp.Kwargs
      description: Keyword arguments for creating the Optuna study.
    - name: optimize_kwargs
      type: tp.Kwargs
      description: Keyword arguments for the optimization process.
  returns:
    - type: optuna.Study
      description: The Optuna study object containing optimization results.
```

--------------------------------

### Get Target (get_target)

Source: https://systematica.mintlify.app/api-reference/systematica/models/meta_model/base

Converts categorical labels into their numeric code representations. This is useful for preparing target variables for machine learning models. The output can be returned as a NumPy array or a Pandas Series.

```APIDOC
get_target(self, to_numpy: bool = False) -> numpy.ndarray | pandas.core.series.Series

Parameters:
  - to_numpy (bool, optional): Output numpy array if True. Pandas DataFrame otherwise. Default to False.

Returns:
  tp.Array1d | pd.Series: Array or Series containing the numeric encoding of categories.
```

--------------------------------

### Systematica Models - Ou Process Module

Source: https://systematica.mintlify.app/api-reference/systematica/models/volatility

Documentation for the Ou Process model within the Systematica project. This section outlines the Ou Process model and its core implementation.

```APIDOC
systematica.models.ou_process

## Sub-modules

* systematica.models.ou_process.ou_process

This module implements the Ornstein-Uhlenbeck (OU) process, a mathematical model often used in finance for modeling interest rates or asset prices. It captures mean reversion properties.
```

--------------------------------

### ColumnStackComposer Constructor and run Method

Source: https://systematica.mintlify.app/api-reference/systematica/portfolio/composers/models

ColumnStackComposer combines trial strategies using the column_stack method from vectorbtpro. The run method executes and combines these strategies using column stacking, with an option to group results.

```APIDOC
ColumnStackComposer:
  Constructor:
    ColumnStackComposer(
        model_name: str,
        signal_params=_Nothing.NOTHING,
        portfolio_params: Dict[str, Any] = _Nothing.NOTHING,
        model_params: Dict[str, Any] = _Nothing.NOTHING,
    )
  Description:
    Combine trial strategies via colum_stack method. A composer that combines multiple trial strategies using the column_stack method from vectorbtpro.
  Ancestors:
    * systematica.portfolio.composers.base.BaseComposer

  run Method:
    Signature:
      run(
          self,
          data: vectorbtpro.data.base.Data,
          s1: str,
          s2: str,
          group_by: bool = True,
          use_close: bool = True,
      ) -> systematica.portfolio.analyzers.base.PortfolioAnalyzer
    Description:
      Execute and combine multiple strategies using column stacking.
    Parameters:
      data (vbt.Data): Input data for the pipeline.
      s1 (str): First symbol.
      s2 (str): Second symbol.
      group_by (bool, optional): Whether to group the results, by default True.
      use_close (bool, optional): Whether to use ‘Close’ prices (default is True).
    Returns:
      PortfolioAnalyzer: Portfolio analyzer object with column-stacked strategy results.
```

--------------------------------

### Plot Cash Report (Python)

Source: https://systematica.mintlify.app/api-reference/systematica/portfolio/analyzers/base

Plots the cash balance of the portfolio over a specified period. It accepts start and end dates and allows for additional plot-specific keyword arguments for customization.

```Python
plot_cash_report(
    self,
    start: str | pandas._libs.tslibs.timestamps.Timestamp = None,
    end: str | pandas._libs.tslibs.timestamps.Timestamp = None,
    **plot_kwargs,
) ‑> vectorbtpro.utils.figure.FigureWidget

Plot the cash balance of the portfolio.

Parameters:

| Name | Type | Default | Description |
| --- | --- | --- | --- |
| `start` | `str` | `None` | The start date. Defaults to `None`. |
| `end` | `str` | `None` | The end date. Defaults to `None`. |
| `plot_kwargs` | `tp.Kwargs` | `--` | Additional plot keyword arguments. |

Returns:

| Type | Description |
| --- | --- |
| `vbt.FigureWidget` | Figure widget of the cash balance. |
```

--------------------------------

### Portfolio Composer Class API

Source: https://systematica.mintlify.app/api-reference/systematica/portfolio/composers/base

Documentation for Systematica portfolio composer classes, covering class variables and the 'run' method. Class variables define model parameters and portfolio simulation arguments. The 'run' method executes the composer with specified data, symbols, and portfolio configurations, returning a PortfolioAnalyzer.

```APIDOC
Class Variables:
  model_name: str
    Name of the model to use. Should match to a model in systematica.api.models.
  portfolio_params: Dict[str, Any]
    Additional keyword arguments for portfolio simulation.
  model_params: Dict[str, Any]
    A dictionary of optional keyword arguments needed to compute the model.

Method: run
  Signature:
    run(
        self,
        data: vectorbtpro.data.base.Data,
        s1: str,
        s2: str,
        portfolio_kwargs: Dict[str, Any],
        **kwargs,
    ) -> systematica.portfolio.analyzers.base.PortfolioAnalyzer
  Description:
    Run the portfolio composer with the specified data.
  Parameters:
    data: vbt.Data
      Vector BT data object containing the input data.
    s1: str
      First symbol or identifier for the data.
    s2: str
      Second symbol or identifier for the data.
    portfolio_kwargs: tp.Kwargs
      Keyword arguments for portfolio configuration.
    kwargs: tp.Kwargs
      Additional keyword arguments passed to the model.
  Returns:
    PortfolioAnalyzer
      Portfolio analyzer instance containing the results of the composition.
```