复制链接

克隆策略

In [76]:

# 本代码由可视化策略环境自动生成 2022年9月19日 12:56
# 本代码单元只能在可视化模式下编辑。您也可以拷贝代码，粘贴到新建的代码单元或者策略，然后修改。


# Python 代码入口函数，input_1/2/3 对应三个输入端，data_1/2/3 对应三个输出端
def m34_run_bigquant_run(input_1, input_2, input_3, SW_type):
    basic_data = input_1.read()
    start_date = str(basic_data.date.min())
    end_date = str(basic_data.date.max())
    SW_data = DataSource('basic_info_IndustrySw').read()
    SW_data['code'] = SW_data.code.astype('int')

    SW_data_2014_2 = SW_data[(SW_data.version==2014)&(SW_data.industry_sw_level==2)][['code','name']].rename(columns={'code':'industry_sw_level2_0','name':'name_SW2'})

    SW_data_2021_2 = SW_data[(SW_data.version==2021)&(SW_data.industry_sw_level==2)][['code','name']].rename(columns={'code':'industry_sw_level2_0','name':'name_SW2'})

    basic_data['daily_return_0'] = basic_data['daily_return_0']-1

    if end_date < '2021-12-13':

        basic_data = basic_data.merge(SW_data_2014_2,how='left',on='industry_sw_level2_0')

    else:
        basic_data_2014 = basic_data[basic_data.date < '2021-12-13']

        basic_data_2014 = basic_data_2014.merge(SW_data_2014_2,how='left',on='industry_sw_level2_0')

        basic_data_2021 = basic_data[basic_data.date >= '2021-12-13']

        basic_data_2021 = basic_data_2021.merge(SW_data_2021_2,how='left',on='industry_sw_level2_0')

        basic_data = pd.concat([basic_data_2014,basic_data_2021])
        
    block_data = basic_data.groupby(['name_SW2','date']).agg({'daily_return_0':'sum','market_cap_float_0':'sum'}).reset_index()
    block_data.columns = ['name_SW2','date','daily_return_0_block_sum','market_cap_float_0_block_sum']
    rst = pd.merge(basic_data[['date','name_SW2','instrument','daily_return_0','market_cap_float_0']],block_data,on=['date','name_SW2'],how='left').dropna()
    
    rst['板块流通收益率'] = rst['daily_return_0']*rst['market_cap_float_0']/rst['market_cap_float_0_block_sum']
    data_1 = DataSource.write_df(rst)
    return Outputs(data_1=data_1)

# 后处理函数，可选。输入是主函数的输出，可以在这里对数据做处理，或者返回更友好的outputs数据格式。此函数输出不会被缓存。
def m34_post_run_bigquant_run(outputs):
    return outputs

# Python 代码入口函数，input_1/2/3 对应三个输入端，data_1/2/3 对应三个输出端
def m44_run_bigquant_run(input_1, input_2, input_3):
    # 示例代码如下。在这里编写您的代码
    from sklearn.model_selection import train_test_split
    data = input_1.read()
    x_train, x_val, y_train, y_val = train_test_split(data["x"], data['y'], test_size=0.2)
    data_1 = DataSource.write_pickle({'x': x_train, 'y': y_train})
    data_2 = DataSource.write_pickle({'x': x_val, 'y': y_val})
    return Outputs(data_1=data_1, data_2=data_2, data_3=None)

# 后处理函数，可选。输入是主函数的输出，可以在这里对数据做处理，或者返回更友好的outputs数据格式。此函数输出不会被缓存。
def m44_post_run_bigquant_run(outputs):
    return outputs

# Python 代码入口函数，input_1/2/3 对应三个输入端，data_1/2/3 对应三个输出端
def m19_run_bigquant_run(input_1, input_2, input_3, SW_type):
    basic_data = input_1.read()
    start_date = str(basic_data.date.min())
    end_date = str(basic_data.date.max())
    SW_data = DataSource('basic_info_IndustrySw').read()
    SW_data['code'] = SW_data.code.astype('int')

    SW_data_2014_2 = SW_data[(SW_data.version==2014)&(SW_data.industry_sw_level==2)][['code','name']].rename(columns={'code':'industry_sw_level2_0','name':'name_SW2'})

    SW_data_2021_2 = SW_data[(SW_data.version==2021)&(SW_data.industry_sw_level==2)][['code','name']].rename(columns={'code':'industry_sw_level2_0','name':'name_SW2'})

    basic_data['daily_return_0'] = basic_data['daily_return_0']-1

    if end_date < '2021-12-13':

        basic_data = basic_data.merge(SW_data_2014_2,how='left',on='industry_sw_level2_0')

    else:
        basic_data_2014 = basic_data[basic_data.date < '2021-12-13']

        basic_data_2014 = basic_data_2014.merge(SW_data_2014_2,how='left',on='industry_sw_level2_0')

        basic_data_2021 = basic_data[basic_data.date >= '2021-12-13']

        basic_data_2021 = basic_data_2021.merge(SW_data_2021_2,how='left',on='industry_sw_level2_0')

        basic_data = pd.concat([basic_data_2014,basic_data_2021])
        
    block_data = basic_data.groupby(['name_SW2','date']).agg({'daily_return_0':'sum','market_cap_float_0':'sum'}).reset_index()
    block_data.columns = ['name_SW2','date','daily_return_0_block_sum','market_cap_float_0_block_sum']
    rst = pd.merge(basic_data[['date','name_SW2','instrument','daily_return_0','market_cap_float_0']],block_data,on=['date','name_SW2'],how='left').dropna()
    
    rst['板块流通收益率'] = rst['daily_return_0']*rst['market_cap_float_0']/rst['market_cap_float_0_block_sum']
    data_1 = DataSource.write_df(rst)
    return Outputs(data_1=data_1)

# 后处理函数，可选。输入是主函数的输出，可以在这里对数据做处理，或者返回更友好的outputs数据格式。此函数输出不会被缓存。
def m19_post_run_bigquant_run(outputs):
    return outputs

import tensorflow as tf
from tensorflow.keras.layers import Layer

def glu(x, n_units=None):
    """Generalized linear unit nonlinear activation."""
    if n_units is None:
        n_units = tf.shape(x)[-1] // 2

    return x[..., :n_units] * tf.nn.sigmoid(x[..., n_units:])


def sparsemax(logits, axis):
    logits = tf.convert_to_tensor(logits, name="logits")

    # We need its original shape for shape inference.
    shape = logits.get_shape()
    rank = shape.rank
    is_last_axis = (axis == -1) or (axis == rank - 1)

    if is_last_axis:
        output = _compute_2d_sparsemax(logits)
        output.set_shape(shape)
        return output

    # Swap logits' dimension of dim and its last dimension.
    rank_op = tf.rank(logits)
    axis_norm = axis % rank
    logits = _swap_axis(logits, axis_norm, tf.math.subtract(rank_op, 1))

    # Do the actual softmax on its last dimension.
    output = _compute_2d_sparsemax(logits)
    output = _swap_axis(output, axis_norm, tf.math.subtract(rank_op, 1))

    # Make shape inference work since transpose may erase its static shape.
    output.set_shape(shape)
    return output


def _swap_axis(logits, dim_index, last_index, **kwargs):
    return tf.transpose(
        logits,
        tf.concat(
            [
                tf.range(dim_index),
                [last_index],
                tf.range(dim_index + 1, last_index),
                [dim_index],
            ],
            0,
        ),
        **kwargs,
    )


def _compute_2d_sparsemax(logits):
    """Performs the sparsemax operation when axis=-1."""
    shape_op = tf.shape(logits)
    obs = tf.math.reduce_prod(shape_op[:-1])
    dims = shape_op[-1]

    # In the paper, they call the logits z.
    # The mean(logits) can be substracted from logits to make the algorithm
    # more numerically stable. the instability in this algorithm comes mostly
    # from the z_cumsum. Substacting the mean will cause z_cumsum to be close
    # to zero. However, in practise the numerical instability issues are very
    # minor and substacting the mean causes extra issues with inf and nan
    # input.
    # Reshape to [obs, dims] as it is almost free and means the remanining
    # code doesn't need to worry about the rank.
    z = tf.reshape(logits, [obs, dims])

    # sort z
    z_sorted, _ = tf.nn.top_k(z, k=dims)

    # calculate k(z)
    z_cumsum = tf.math.cumsum(z_sorted, axis=-1)
    k = tf.range(1, tf.cast(dims, logits.dtype) + 1) #, dtype=logits.dtype)
    z_check = 1 + k * z_sorted > z_cumsum
    # because the z_check vector is always [1,1,...1,0,0,...0] finding the
    # (index + 1) of the last `1` is the same as just summing the number of 1.
    k_z = tf.math.reduce_sum(tf.cast(z_check, tf.int32), axis=-1)

    # calculate tau(z)
    # If there are inf values or all values are -inf, the k_z will be zero,
    # this is mathematically invalid and will also cause the gather_nd to fail.
    # Prevent this issue for now by setting k_z = 1 if k_z = 0, this is then
    # fixed later (see p_safe) by returning p = nan. This results in the same
    # behavior as softmax.
    k_z_safe = tf.math.maximum(k_z, 1)
    indices = tf.stack([tf.range(0, obs), tf.reshape(k_z_safe, [-1]) - 1], axis=1)
    tau_sum = tf.gather_nd(z_cumsum, indices)
    tau_z = (tau_sum - 1) / tf.cast(k_z, logits.dtype)

    # calculate p
    p = tf.math.maximum(tf.cast(0, logits.dtype), z - tf.expand_dims(tau_z, -1))
    # If k_z = 0 or if z = nan, then the input is invalid
    p_safe = tf.where(
        tf.expand_dims(
            tf.math.logical_or(tf.math.equal(k_z, 0), tf.math.is_nan(z_cumsum[:, -1])),
            axis=-1,
        ),
        tf.fill([obs, dims], tf.cast(float("nan"), logits.dtype)),
        p,
    )

    # Reshape back to original size
    p_safe = tf.reshape(p_safe, shape_op)
    return p_safe


"""
Code replicated from https://github.com/tensorflow/addons/blob/master/tensorflow_addons/layers/normalizations.py
"""
class GroupNormalization(tf.keras.layers.Layer):
    def __init__(
            self,
            groups: int = 2,
            axis: int = -1,
            epsilon: float = 1e-3,
            center: bool = True,
            scale: bool = True,
            beta_initializer="zeros",
            gamma_initializer="ones",
            beta_regularizer=None,
            gamma_regularizer=None,
            beta_constraint=None,
            gamma_constraint=None,
            **kwargs
    ):
        super().__init__(**kwargs)
        self.supports_masking = True
        self.groups = groups
        self.axis = axis
        self.epsilon = epsilon
        self.center = center
        self.scale = scale
        self.beta_initializer = tf.keras.initializers.get(beta_initializer)
        self.gamma_initializer = tf.keras.initializers.get(gamma_initializer)
        self.beta_regularizer = tf.keras.regularizers.get(beta_regularizer)
        self.gamma_regularizer = tf.keras.regularizers.get(gamma_regularizer)
        self.beta_constraint = tf.keras.constraints.get(beta_constraint)
        self.gamma_constraint = tf.keras.constraints.get(gamma_constraint)
        self._check_axis()

    def build(self, input_shape):

        self._check_if_input_shape_is_none(input_shape)
        self._set_number_of_groups_for_instance_norm(input_shape)
        self._check_size_of_dimensions(input_shape)
        self._create_input_spec(input_shape)

        self._add_gamma_weight(input_shape)
        self._add_beta_weight(input_shape)
        self.built = True
        super().build(input_shape)

    def call(self, inputs, training=None):
        # Training=none is just for compat with batchnorm signature call
        input_shape = tf.keras.backend.int_shape(inputs)
        tensor_input_shape = tf.shape(inputs)

        reshaped_inputs, group_shape = self._reshape_into_groups(
            inputs, input_shape, tensor_input_shape
        )

        normalized_inputs = self._apply_normalization(reshaped_inputs, input_shape)

        outputs = tf.reshape(normalized_inputs, tensor_input_shape)

        return outputs

    def get_config(self):
        config = {
            "groups": self.groups,
            "axis": self.axis,
            "epsilon": self.epsilon,
            "center": self.center,
            "scale": self.scale,
            "beta_initializer": tf.keras.initializers.serialize(self.beta_initializer),
            "gamma_initializer": tf.keras.initializers.serialize(
                self.gamma_initializer
            ),
            "beta_regularizer": tf.keras.regularizers.serialize(self.beta_regularizer),
            "gamma_regularizer": tf.keras.regularizers.serialize(
                self.gamma_regularizer
            ),
            "beta_constraint": tf.keras.constraints.serialize(self.beta_constraint),
            "gamma_constraint": tf.keras.constraints.serialize(self.gamma_constraint),
        }
        base_config = super().get_config()
        return {**base_config, **config}

    def compute_output_shape(self, input_shape):
        return input_shape

    def _reshape_into_groups(self, inputs, input_shape, tensor_input_shape):

        group_shape = [tensor_input_shape[i] for i in range(len(input_shape))]
        group_shape[self.axis] = input_shape[self.axis] // self.groups
        group_shape.insert(self.axis, self.groups)
        group_shape = tf.stack(group_shape)
        reshaped_inputs = tf.reshape(inputs, group_shape)
        return reshaped_inputs, group_shape

    def _apply_normalization(self, reshaped_inputs, input_shape):

        group_shape = tf.keras.backend.int_shape(reshaped_inputs)
        group_reduction_axes = list(range(1, len(group_shape)))
        axis = -2 if self.axis == -1 else self.axis - 1
        group_reduction_axes.pop(axis)

        mean, variance = tf.nn.moments(
            reshaped_inputs, group_reduction_axes, keepdims=True
        )

        gamma, beta = self._get_reshaped_weights(input_shape)
        normalized_inputs = tf.nn.batch_normalization(
            reshaped_inputs,
            mean=mean,
            variance=variance,
            scale=gamma,
            offset=beta,
            variance_epsilon=self.epsilon,
        )
        return normalized_inputs

    def _get_reshaped_weights(self, input_shape):
        broadcast_shape = self._create_broadcast_shape(input_shape)
        gamma = None
        beta = None
        if self.scale:
            gamma = tf.reshape(self.gamma, broadcast_shape)

        if self.center:
            beta = tf.reshape(self.beta, broadcast_shape)
        return gamma, beta

    def _check_if_input_shape_is_none(self, input_shape):
        dim = input_shape[self.axis]
        if dim is None:
            raise ValueError(
                "Axis " + str(self.axis) + " of "
                                           "input tensor should have a defined dimension "
                                           "but the layer received an input with shape " + str(input_shape) + "."
            )

    def _set_number_of_groups_for_instance_norm(self, input_shape):
        dim = input_shape[self.axis]

        if self.groups == -1:
            self.groups = dim

    def _check_size_of_dimensions(self, input_shape):

        dim = input_shape[self.axis]
        if dim < self.groups:
            raise ValueError(
                "Number of groups (" + str(self.groups) + ") cannot be "
                                                          "more than the number of channels (" + str(dim) + ")."
            )

        if dim % self.groups != 0:
            raise ValueError(
                "Number of groups (" + str(self.groups) + ") must be a "
                                                          "multiple of the number of channels (" + str(dim) + ")."
            )

    def _check_axis(self):

        if self.axis == 0:
            raise ValueError(
                "You are trying to normalize your batch axis. Do you want to "
                "use tf.layer.batch_normalization instead"
            )

    def _create_input_spec(self, input_shape):

        dim = input_shape[self.axis]
        self.input_spec = tf.keras.layers.InputSpec(
            ndim=len(input_shape), axes={self.axis: dim}
        )

    def _add_gamma_weight(self, input_shape):

        dim = input_shape[self.axis]
        shape = (dim,)

        if self.scale:
            self.gamma = self.add_weight(
                shape=shape,
                name="gamma",
                initializer=self.gamma_initializer,
                regularizer=self.gamma_regularizer,
                constraint=self.gamma_constraint,
            )
        else:
            self.gamma = None

    def _add_beta_weight(self, input_shape):

        dim = input_shape[self.axis]
        shape = (dim,)

        if self.center:
            self.beta = self.add_weight(
                shape=shape,
                name="beta",
                initializer=self.beta_initializer,
                regularizer=self.beta_regularizer,
                constraint=self.beta_constraint,
            )
        else:
            self.beta = None

    def _create_broadcast_shape(self, input_shape):
        broadcast_shape = [1] * len(input_shape)
        broadcast_shape[self.axis] = input_shape[self.axis] // self.groups
        broadcast_shape.insert(self.axis, self.groups)
        return broadcast_shape

    
class TransformBlock(tf.keras.layers.Layer):

    def __init__(self, features,
                 norm_type,
                 momentum=0.9,
                 virtual_batch_size=None,
                 groups=2,
                 block_name='',
                 **kwargs):
        super(TransformBlock, self).__init__(**kwargs)

        self.features = features
        self.norm_type = norm_type
        self.momentum = momentum
        self.groups = groups
        self.virtual_batch_size = virtual_batch_size
        self.block_name = block_name
    
    def build(self, input_shape):
        self.transform = tf.keras.layers.Dense(self.features, use_bias=False, name=f'transformblock_dense_{self.block_name}')
        if self.norm_type == 'batch':
            self.bn = tf.keras.layers.BatchNormalization(axis=-1, momentum=momentum,
                                                         virtual_batch_size=virtual_batch_size,
                                                         name=f'transformblock_bn_{self.block_name}')
        else:
            self.bn = GroupNormalization(axis=-1, groups=self.groups, name=f'transformblock_gn_{self.block_name}')
            
        self.built = True
        super().build(input_shape)
        
    def call(self, inputs, training=None):
        x = self.transform(inputs)
        x = self.bn(x, training=training)
        return x
    
    def get_config(self):
        config = {
            "features": self.features,
            "norm_type": self.norm_type,
            "virtual_batch_size": self.virtual_batch_size,
            "groups": self.groups,
            "block_name": self.block_name
        }
        base_config = super().get_config()
        return {**base_config, **config}
    
    def compute_output_shape(self, input_shape):
        return input_shape


class TabNetEncoderLayer(tf.keras.layers.Layer):

    def __init__(self, feature_columns,
                 feature_dim=16,
                 output_dim=8,
                 num_features=None,
                 num_decision_steps=3,
                 relaxation_factor=1.5,
                 sparsity_coefficient=1e-5,
                 norm_type='group',
                 batch_momentum=0.98,
                 virtual_batch_size=1024,
                 num_groups=2,
                 epsilon=1e-5,
                 **kwargs):

        super(TabNetEncoderLayer, self).__init__(**kwargs)

        # Input checks
        if feature_columns is not None:
            if type(feature_columns) not in (list, tuple):
                raise ValueError("`feature_columns` must be a list or a tuple.")

            if len(feature_columns) == 0:
                raise ValueError("`feature_columns` must be contain at least 1 tf.feature_column !")

            if num_features is None:
                num_features = len(feature_columns)
            else:
                num_features = int(num_features)

        else:
            if num_features is None:
                raise ValueError("If `feature_columns` is None, then `num_features` cannot be None.")

        if num_decision_steps < 1:
            raise ValueError("Num decision steps must be greater than 0.")
        
        if feature_dim <= output_dim:
            raise ValueError("To compute `features_for_coef`, feature_dim must be larger than output dim")

        feature_dim = int(feature_dim)
        output_dim = int(output_dim)
        num_decision_steps = int(num_decision_steps)
        relaxation_factor = float(relaxation_factor)
        sparsity_coefficient = float(sparsity_coefficient)
        batch_momentum = float(batch_momentum)
        num_groups = max(1, int(num_groups))
        epsilon = float(epsilon)

        if relaxation_factor < 0.:
            raise ValueError("`relaxation_factor` cannot be negative !")

        if sparsity_coefficient < 0.:
            raise ValueError("`sparsity_coefficient` cannot be negative !")

        if virtual_batch_size is not None:
            virtual_batch_size = int(virtual_batch_size)

        if norm_type not in ['batch', 'group']:
            raise ValueError("`norm_type` must be either `batch` or `group`")

        self.feature_columns = feature_columns
        self.num_features = num_features
        self.feature_dim = feature_dim
        self.output_dim = output_dim

        self.num_decision_steps = num_decision_steps
        self.relaxation_factor = relaxation_factor
        self.sparsity_coefficient = sparsity_coefficient
        self.norm_type = norm_type
        self.batch_momentum = batch_momentum
        self.virtual_batch_size = virtual_batch_size
        self.num_groups = num_groups
        self.epsilon = epsilon

        if num_decision_steps > 1:
            features_for_coeff = feature_dim - output_dim
            print(f"[TabNet]: {features_for_coeff} features will be used for decision steps.")

        if self.feature_columns is not None:
            self.input_features = tf.keras.layers.DenseFeatures(feature_columns, trainable=True)

            if self.norm_type == 'batch':
                self.input_bn = tf.keras.layers.BatchNormalization(axis=-1, momentum=batch_momentum, name='input_bn')
            else:
                self.input_bn = GroupNormalization(axis=-1, groups=self.num_groups, name='input_gn')

        else:
            self.input_features = None
            self.input_bn = None
    
    def build(self, input_shape):
        self.transform_f1 = TransformBlock(2 * self.feature_dim, self.norm_type,
                                           self.batch_momentum, self.virtual_batch_size, self.num_groups,
                                           block_name='f1')

        self.transform_f2 = TransformBlock(2 * self.feature_dim, self.norm_type,
                                           self.batch_momentum, self.virtual_batch_size, self.num_groups,
                                           block_name='f2')

        self.transform_f3_list = [
            TransformBlock(2 * self.feature_dim, self.norm_type,
                           self.batch_momentum, self.virtual_batch_size, self.num_groups, block_name=f'f3_{i}')
            for i in range(self.num_decision_steps)
        ]

        self.transform_f4_list = [
            TransformBlock(2 * self.feature_dim, self.norm_type,
                           self.batch_momentum, self.virtual_batch_size, self.num_groups, block_name=f'f4_{i}')
            for i in range(self.num_decision_steps)
        ]

        self.transform_coef_list = [
            TransformBlock(self.num_features, self.norm_type,
                           self.batch_momentum, self.virtual_batch_size, self.num_groups, block_name=f'coef_{i}')
            for i in range(self.num_decision_steps - 1)
        ]

        self._step_feature_selection_masks = None
        self._step_aggregate_feature_selection_mask = None
        self.built = True
        super(TabNetEncoderLayer, self).build(input_shape)

    def call(self, inputs, training=None):
        if self.input_features is not None:
            features = self.input_features(inputs)
            features = self.input_bn(features, training=training)

        else:
            features = inputs

        batch_size = tf.shape(features)[0]
        self._step_feature_selection_masks = []
        self._step_aggregate_feature_selection_mask = None

        # Initializes decision-step dependent variables.
        output_aggregated = tf.zeros([batch_size, self.output_dim])
        masked_features = features
        mask_values = tf.zeros([batch_size, self.num_features])
        aggregated_mask_values = tf.zeros([batch_size, self.num_features])
        complementary_aggregated_mask_values = tf.ones(
            [batch_size, self.num_features])

        total_entropy = 0.0
        entropy_loss = 0.

        for ni in range(self.num_decision_steps):
            # Feature transformer with two shared and two decision step dependent
            # blocks is used below.=
            transform_f1 = self.transform_f1(masked_features, training=training)
            transform_f1 = glu(transform_f1, self.feature_dim)

            transform_f2 = self.transform_f2(transform_f1, training=training)
            transform_f2 = (glu(transform_f2, self.feature_dim) +
                            transform_f1) * tf.math.sqrt(0.5)

            transform_f3 = self.transform_f3_list[ni](transform_f2, training=training)
            transform_f3 = (glu(transform_f3, self.feature_dim) +
                            transform_f2) * tf.math.sqrt(0.5)

            transform_f4 = self.transform_f4_list[ni](transform_f3, training=training)
            transform_f4 = (glu(transform_f4, self.feature_dim) +
                            transform_f3) * tf.math.sqrt(0.5)

            if (ni > 0 or self.num_decision_steps == 1):
                decision_out = tf.nn.relu(transform_f4[:, :self.output_dim])

                # Decision aggregation.
                output_aggregated += decision_out

                # Aggregated masks are used for visualization of the
                # feature importance attributes.
                scale_agg = tf.reduce_sum(decision_out, axis=1, keepdims=True)

                if self.num_decision_steps > 1:
                    scale_agg = scale_agg / tf.cast(self.num_decision_steps - 1, tf.float32)

                aggregated_mask_values += mask_values * scale_agg

            features_for_coef = transform_f4[:, self.output_dim:]

            if ni < (self.num_decision_steps - 1):
                # Determines the feature masks via linear and nonlinear
                # transformations, taking into account of aggregated feature use.
                mask_values = self.transform_coef_list[ni](features_for_coef, training=training)
                mask_values *= complementary_aggregated_mask_values
                mask_values = sparsemax(mask_values, axis=-1)

                # Relaxation factor controls the amount of reuse of features between
                # different decision blocks and updated with the values of
                # coefficients.
                complementary_aggregated_mask_values *= (
                        self.relaxation_factor - mask_values)

                # Entropy is used to penalize the amount of sparsity in feature
                # selection.
                total_entropy += tf.reduce_mean(
                    tf.reduce_sum(
                        -mask_values * tf.math.log(mask_values + self.epsilon), axis=1)) / (
                                     tf.cast(self.num_decision_steps - 1, tf.float32))

                # Add entropy loss
                entropy_loss = total_entropy

                # Feature selection.
                masked_features = tf.multiply(mask_values, features)

                # Visualization of the feature selection mask at decision step ni
                # tf.summary.image(
                #     "Mask for step" + str(ni),
                #     tf.expand_dims(tf.expand_dims(mask_values, 0), 3),
                #     max_outputs=1)
                mask_at_step_i = tf.expand_dims(tf.expand_dims(mask_values, 0), 3)
                self._step_feature_selection_masks.append(mask_at_step_i)

            else:
                # This branch is needed for correct compilation by tf.autograph
                entropy_loss = 0.

        # Adds the loss automatically
        self.add_loss(self.sparsity_coefficient * entropy_loss)

        # Visualization of the aggregated feature importances
        # tf.summary.image(
        #     "Aggregated mask",
        #     tf.expand_dims(tf.expand_dims(aggregated_mask_values, 0), 3),
        #     max_outputs=1)

        agg_mask = tf.expand_dims(tf.expand_dims(aggregated_mask_values, 0), 3)
        self._step_aggregate_feature_selection_mask = agg_mask
        return output_aggregated

    def feature_selection_masks(self):
        return self._step_feature_selection_masks

    def aggregate_feature_selection_mask(self):
        return self._step_aggregate_feature_selection_mask
    
    def compute_output_shape(self, input_shape):
        return self.output_dim
    
    def get_config(self):
        config = {
            "feature_columns": self.feature_columns,
            "num_features": self.num_features,
            "feature_dim": self.feature_dim,
            "output_dim": self.output_dim,
            "num_decision_steps": self.num_decision_steps,
            "relaxation_factor": self.relaxation_factor,
            "sparsity_coefficient": self.sparsity_coefficient,
            "norm_type": self.norm_type,
            "batch_momentum": self.batch_momentum,
            "virtual_batch_size": self.virtual_batch_size,
            "num_groups": self.num_groups,
            "epsilon": self.epsilon,
        }
        base_config = super().get_config()
        return {**base_config, **config}
    
    
# 必须也将 UserLayer 赋值给 m38_layer_class_bigquant_run
m38_layer_class_bigquant_run = TabNetEncoderLayer

# 用户的自定义层需要写到字典中，比如
# {
#   "MyLayer": MyLayer
# }
m45_custom_objects_bigquant_run = {
    "GroupNormalization": GroupNormalization,
    "TransformBlock": TransformBlock,
    "TabNetEncoderLayer": TabNetEncoderLayer
}

# Python 代码入口函数，input_1/2/3 对应三个输入端，data_1/2/3 对应三个输出端
def m10_run_bigquant_run(input_1, input_2, input_3):
    # 示例代码如下。在这里编写您的代码
    pred_label = input_1.read_pickle()
    df = input_2.read_df()
    df = pd.DataFrame({'pred_label':pred_label[:,0], 'instrument':df.instrument, 'date':df.date})
    df.sort_values(['date','pred_label'],inplace=True, ascending=[True,False])
    return Outputs(data_1=DataSource.write_df(df), data_2=None, data_3=None)

# 后处理函数，可选。输入是主函数的输出，可以在这里对数据做处理，或者返回更友好的outputs数据格式。此函数输出不会被缓存。
def m10_post_run_bigquant_run(outputs):
    return outputs

# 回测引擎：初始化函数，只执行一次
def m6_initialize_bigquant_run(context):
    # 加载预测数据
    context.ranker_prediction = context.options['data'].read_df()

    # 系统已经设置了默认的交易手续费和滑点，要修改手续费可使用如下函数
    context.set_commission(PerOrder(buy_cost=0.001, sell_cost=0.001, min_cost=5))
    # 预测数据，通过options传入进来，使用 read_df 函数，加载到内存 (DataFrame)
    # 设置买入的股票数量，这里买入预测股票列表排名靠前的5只
    stock_count = 1
    # 每只的股票的权重，如下的权重分配会使得靠前的股票分配多一点的资金，[0.339160, 0.213986, 0.169580, ..]
    context.stock_weights = T.norm([1 / math.log(i + 2) for i in range(0, stock_count)])
    # 设置每只股票占用的最大资金比例
    context.max_cash_per_instrument = 0.2
    context.options['hold_days'] = 2

# 回测引擎：每日数据处理函数，每天执行一次
def m6_handle_data_bigquant_run(context, data):
    # 按日期过滤得到今日的预测数据
    ranker_prediction = context.ranker_prediction[
        context.ranker_prediction.date == data.current_dt.strftime('%Y-%m-%d')]

    # 1. 资金分配
    # 平均持仓时间是hold_days，每日都将买入股票，每日预期使用 1/hold_days 的资金
    # 实际操作中，会存在一定的买入误差，所以在前hold_days天，等量使用资金；之后，尽量使用剩余资金（这里设置最多用等量的1.5倍）
    is_staging = context.trading_day_index < context.options['hold_days'] # 是否在建仓期间（前 hold_days 天）
    cash_avg = context.portfolio.portfolio_value / context.options['hold_days']
    cash_for_buy = min(context.portfolio.cash, (1 if is_staging else 1.5) * cash_avg)
    cash_for_sell = cash_avg - (context.portfolio.cash - cash_for_buy)
    positions = {e.symbol: p.amount * p.last_sale_price
                 for e, p in context.perf_tracker.position_tracker.positions.items()}

    # 2. 生成卖出订单：hold_days天之后才开始卖出；对持仓的股票，按机器学习算法预测的排序末位淘汰
    if not is_staging and cash_for_sell > 0:
        equities = {e.symbol: e for e, p in context.perf_tracker.position_tracker.positions.items()}
        instruments = list(reversed(list(ranker_prediction.instrument[ranker_prediction.instrument.apply(
                lambda x: x in equities and not context.has_unfinished_sell_order(equities[x]))])))
        # print('rank order for sell %s' % instruments)
        for instrument in instruments:
            context.order_target(context.symbol(instrument), 0)
            cash_for_sell -= positions[instrument]
            if cash_for_sell <= 0:
                break

    # 3. 生成买入订单：按机器学习算法预测的排序，买入前面的stock_count只股票
    buy_cash_weights = context.stock_weights
    buy_instruments = list(ranker_prediction.instrument[:len(buy_cash_weights)])
    max_cash_per_instrument = context.portfolio.portfolio_value * context.max_cash_per_instrument
    for i, instrument in enumerate(buy_instruments):
        cash = cash_for_buy * buy_cash_weights[i]
        if cash > max_cash_per_instrument - positions.get(instrument, 0):
            # 确保股票持仓量不会超过每次股票最大的占用资金量
            cash = max_cash_per_instrument - positions.get(instrument, 0)
        if cash > 0:
            context.order_value(context.symbol(instrument), cash)

# 回测引擎：准备数据，只执行一次
def m6_prepare_bigquant_run(context):
    pass


m1 = M.instruments.v2(
    start_date='2019-06-01',
    end_date='2020-01-01',
    market='CN_STOCK_A',
    instrument_list='',
    max_count=0
)

m20 = M.use_datasource.v1(
    instruments=m1.data,
    datasource_id='bar1d_CN_STOCK_A',
    start_date='',
    end_date=''
)

m9 = M.instruments.v2(
    start_date=T.live_run_param('trading_date', '2020-01-01'),
    end_date=T.live_run_param('trading_date', '2021-06-22'),
    market='CN_STOCK_A',
    instrument_list='',
    max_count=0
)

m5 = M.use_datasource.v1(
    instruments=m9.data,
    datasource_id='bar1d_CN_STOCK_A',
    start_date='',
    end_date=''
)

m2 = M.input_features.v1(
    features="""# #号开始的表示注释
# 多个特征，每行一个，可以包含基础特征和衍生特征

close_0/mean(close_0,5)
close_0/mean(close_0,10)
close_0/mean(close_0,20)
close_0/open_0
open_0/mean(close_0,5)
open_0/mean(close_0,10)
open_0/mean(close_0,20)
 
close_1
open_1
high_1
low_1
return_1
amount_1
turn_1
 
close_2
open_2
high_2
low_2
amount_2
turn_2
return_2
 
close_3
open_3
high_3
low_3
amount_3
turn_3
return_3
 
close_4
open_4
high_4
low_4
amount_4
turn_4
return_4
 
mean(close_0, 2)
mean(low_0, 2)
mean(open_0, 2)
mean(high_0, 2)
mean(turn_0, 2)
mean(amount_0, 2)
mean(return_0, 2)
 
ts_max(close_0, 2)
ts_max(low_0, 2)
ts_max(open_0, 2)
ts_max(high_0, 2)
ts_max(turn_0, 2)
ts_max(amount_0, 2)
ts_max(return_0, 2)
 
ts_min(close_0, 2)
ts_min(low_0, 2)
ts_min(open_0, 2)
ts_min(high_0, 2)
ts_min(turn_0, 2)
ts_min(amount_0, 2)
ts_min(return_0, 2) 
 
std(close_0, 2)
std(low_0, 2)
std(open_0, 2)
std(high_0, 2)
std(turn_0, 2)
std(amount_0, 2)
std(return_0, 2)
 
ts_rank(close_0, 2)
ts_rank(low_0, 2)
ts_rank(open_0, 2)
ts_rank(high_0, 2)
ts_rank(turn_0, 2)
ts_rank(amount_0, 2)
ts_rank(return_0, 2)

decay_linear(close_0, 2)
decay_linear(low_0, 2)
decay_linear(open_0, 2)
decay_linear(high_0, 2)
decay_linear(turn_0, 2)
decay_linear(amount_0, 2)
decay_linear(return_0, 2)
 
correlation(volume_0, return_0, 2)
correlation(volume_0, high_0, 2)
correlation(volume_0, low_0, 2)
correlation(volume_0, close_0, 2)
correlation(volume_0, open_0, 2)
correlation(volume_0, turn_0, 2)
  
correlation(return_0, high_0, 2)
correlation(return_0, low_0, 2)
correlation(return_0, close_0, 2)
correlation(return_0, open_0, 2)
correlation(return_0, turn_0, 2)
 
correlation(high_0, low_0, 2)
correlation(high_0, close_0, 2)
correlation(high_0, open_0, 2)
correlation(high_0, turn_0, 2)
 
correlation(low_0, close_0, 2)
correlation(low_0, open_0, 2)
correlation(low_0, turn_0, 2)
 
correlation(close_0, open_0, 2)
correlation(close_0, turn_0, 2)

correlation(open_0, turn_0, 2)
"""
)

m15 = M.general_feature_extractor.v7(
    instruments=m1.data,
    features=m2.data,
    start_date='',
    end_date='',
    before_start_days=90
)

m16 = M.derived_feature_extractor.v3(
    input_data=m15.data,
    features=m2.data,
    date_col='date',
    instrument_col='instrument',
    drop_na=False,
    remove_extra_columns=False
)

m17 = M.general_feature_extractor.v7(
    instruments=m9.data,
    features=m2.data,
    start_date='',
    end_date='',
    before_start_days=90
)

m18 = M.derived_feature_extractor.v3(
    input_data=m17.data,
    features=m2.data,
    date_col='date',
    instrument_col='instrument',
    drop_na=False,
    remove_extra_columns=False
)

m26 = M.filter_stockcode.v2(
    input_1=m18.data,
    start='688'
)

m27 = M.filtet_st_stock.v7(
    input_1=m26.data_1
)

m28 = M.filter_delist_stocks.v3(
    input_1=m27.data_1
)

m41 = M.standardlize.v8(
    input_1=m16.data,
    input_2=m2.data,
    columns_input='[]'
)

m42 = M.fillnan.v1(
    input_data=m41.data,
    features=m2.data,
    fill_value='0.0'
)

m31 = M.input_features.v1(
    features="""industry_sw_level2_0
market_cap_float_0
daily_return_0
"""
)

m33 = M.general_feature_extractor.v7(
    instruments=m1.data,
    features=m31.data,
    start_date='',
    end_date='',
    before_start_days=90
)

m34 = M.cached.v3(
    input_1=m33.data,
    run=m34_run_bigquant_run,
    post_run=m34_post_run_bigquant_run,
    input_ports='',
    params="""{
    'SW_type':'name_SW2'
}""",
    output_ports='data_1,data_2'
)

m12 = M.join.v3(
    data1=m34.data_1,
    data2=m20.data,
    on='date,instrument',
    how='inner',
    sort=False
)

m25 = M.filter.v3(
    input_data=m12.data,
    expr='(rank(sum(板块流通收益率,5))>=0.6)',
    output_left_data=False
)

m32 = M.filter.v3(
    input_data=m25.data,
    expr='(rank(sum(板块流通收益率,5))>=0.7)',
    output_left_data=False
)

m21 = M.auto_labeler_on_datasource.v1(
    input_data=m32.data,
    label_expr="""

# 计算收益：5日收盘价(作为卖出价格)除以明日开盘价(作为买入价格)
shift(close, -5) / shift(open, -1)-1
# 极值处理：用1%和99%分位的值做clip
clip(label, all_quantile(label, 0.01), all_quantile(label, 0.99))

# 将分数映射到分类，这里使用20个分类
all_wbins(label, 2)

# 过滤掉一字涨停的情况 (设置label为NaN，在后续处理和训练中会忽略NaN的label)
where(shift(high, -1) == shift(low, -1), NaN, label)""",
    drop_na_label=True,
    cast_label_int=True,
    date_col='date',
    instrument_col='instrument',
    user_functions={}
)

m40 = M.standardlize.v8(
    input_1=m21.data,
    columns_input='label'
)

m7 = M.join.v3(
    data1=m40.data,
    data2=m42.data,
    on='date,instrument',
    how='inner',
    sort=False
)

m43 = M.dl_convert_to_bin.v2(
    input_data=m7.data,
    features=m2.data,
    window_size=1,
    feature_clip=3,
    flatten=False,
    window_along_col='instrument'
)

m44 = M.cached.v3(
    input_1=m43.data,
    run=m44_run_bigquant_run,
    post_run=m44_post_run_bigquant_run,
    input_ports='',
    params='{}',
    output_ports=''
)

m4 = M.input_features.v1(
    features="""industry_sw_level2_0
market_cap_float_0
daily_return_0
"""
)

m11 = M.general_feature_extractor.v7(
    instruments=m9.data,
    features=m4.data,
    start_date='',
    end_date='',
    before_start_days=90
)

m19 = M.cached.v3(
    input_1=m11.data,
    run=m19_run_bigquant_run,
    post_run=m19_post_run_bigquant_run,
    input_ports='',
    params="""{
    'SW_type':'name_SW2'
}""",
    output_ports='data_1,data_2'
)

m22 = M.join.v3(
    data1=m19.data_1,
    data2=m5.data,
    on='date,instrument',
    how='inner',
    sort=False
)

m24 = M.join.v3(
    data1=m28.data,
    data2=m22.data,
    on='date,instrument',
    how='inner',
    sort=False
)

m35 = M.filter.v3(
    input_data=m24.data,
    expr='(rank(sum(板块流通收益率,5))>=0.7)',
    output_left_data=False
)

m36 = M.filter.v3(
    input_data=m35.data,
    expr='(rank(sum(板块流通收益率,5))>=0.7)',
    output_left_data=False
)

m14 = M.dropnan.v1(
    input_data=m36.data
)

m39 = M.dl_convert_to_bin.v2(
    input_data=m14.data,
    features=m2.data,
    window_size=1,
    feature_clip=3,
    flatten=False,
    window_along_col='instrument'
)

m8 = M.join.v3(
    data1=m14.data,
    data2=m18.data,
    on='date,instrument',
    how='inner',
    sort=False
)

m3 = M.dl_layer_input.v1(
    shape='98',
    batch_shape='',
    dtype='float32',
    sparse=False,
    name=''
)

m38 = M.dl_layer_userlayer.v1(
    input1=m3.data,
    layer_class=m38_layer_class_bigquant_run,
    params="""{
    "num_features": 98, 
    "feature_columns": None,
    "feature_dim": 64,
    "output_dim": 32,
    "num_decision_steps": 3,
    "relaxation_factor": 1.3,
    "sparsity_coefficient": 1e-5,
    "norm_type": "group",
    "batch_momentum": 0.9,
    "virtual_batch_size": 128,
    "num_groups": 2,
    "epsilon": 1e-5
}""",
    name=''
)

m23 = M.dl_layer_dense.v1(
    inputs=m38.data,
    units=1,
    activation='linear',
    use_bias=False,
    kernel_initializer='Zeros',
    bias_initializer='Zeros',
    kernel_regularizer='None',
    kernel_regularizer_l1=0,
    kernel_regularizer_l2=0,
    bias_regularizer='None',
    bias_regularizer_l1=0,
    bias_regularizer_l2=0,
    activity_regularizer='None',
    activity_regularizer_l1=0,
    activity_regularizer_l2=0,
    kernel_constraint='None',
    bias_constraint='None',
    name=''
)

m37 = M.dl_model_init.v1(
    inputs=m3.data,
    outputs=m23.data
)

m45 = M.dl_model_train.v1(
    input_model=m37.data,
    training_data=m44.data_1,
    validation_data=m44.data_2,
    optimizer='Adam',
    loss='mean_squared_error',
    metrics='mse',
    batch_size=10240,
    epochs=10,
    custom_objects=m45_custom_objects_bigquant_run,
    n_gpus=0,
    verbose='2:每个epoch输出一行记录',
    m_cached=False
)

m47 = M.dl_model_predict.v1(
    trained_model=m45.data,
    input_data=m39.data,
    batch_size=1024,
    n_gpus=0,
    verbose='2:每个epoch输出一行记录'
)

m10 = M.cached.v3(
    input_1=m47.data,
    input_2=m8.data,
    run=m10_run_bigquant_run,
    post_run=m10_post_run_bigquant_run,
    input_ports='',
    params='{}',
    output_ports=''
)

m6 = M.trade.v4(
    instruments=m2.data,
    options_data=m10.data_1,
    start_date='',
    end_date='',
    initialize=m6_initialize_bigquant_run,
    handle_data=m6_handle_data_bigquant_run,
    prepare=m6_prepare_bigquant_run,
    volume_limit=0.025,
    order_price_field_buy='open',
    order_price_field_sell='close',
    capital_base=1000000,
    auto_cancel_non_tradable_orders=True,
    data_frequency='daily',
    price_type='后复权',
    product_type='股票',
    plot_charts=True,
    backtest_only=False,
    benchmark='000300.SHA'
)

[2022-09-19 12:43:35.157842] INFO: moduleinvoker: instruments.v2 开始运行..

[2022-09-19 12:43:35.171413] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.172963] INFO: moduleinvoker: instruments.v2 运行完成[0.015127s].

[2022-09-19 12:43:35.180920] INFO: moduleinvoker: use_datasource.v1 开始运行..

[2022-09-19 12:43:35.189789] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.191428] INFO: moduleinvoker: use_datasource.v1 运行完成[0.010524s].

[2022-09-19 12:43:35.196403] INFO: moduleinvoker: instruments.v2 开始运行..

[2022-09-19 12:43:35.201876] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.203770] INFO: moduleinvoker: instruments.v2 运行完成[0.007365s].

[2022-09-19 12:43:35.209734] INFO: moduleinvoker: use_datasource.v1 开始运行..

[2022-09-19 12:43:35.215595] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.217592] INFO: moduleinvoker: use_datasource.v1 运行完成[0.007863s].

[2022-09-19 12:43:35.222631] INFO: moduleinvoker: input_features.v1 开始运行..

[2022-09-19 12:43:35.232658] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.234520] INFO: moduleinvoker: input_features.v1 运行完成[0.011894s].

[2022-09-19 12:43:35.249980] INFO: moduleinvoker: general_feature_extractor.v7 开始运行..

[2022-09-19 12:43:35.256530] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.258374] INFO: moduleinvoker: general_feature_extractor.v7 运行完成[0.008418s].

[2022-09-19 12:43:35.264775] INFO: moduleinvoker: derived_feature_extractor.v3 开始运行..

[2022-09-19 12:43:35.282838] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.284671] INFO: moduleinvoker: derived_feature_extractor.v3 运行完成[0.019895s].

[2022-09-19 12:43:35.297834] INFO: moduleinvoker: general_feature_extractor.v7 开始运行..

[2022-09-19 12:43:35.303396] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.305087] INFO: moduleinvoker: general_feature_extractor.v7 运行完成[0.007271s].

[2022-09-19 12:43:35.312021] INFO: moduleinvoker: derived_feature_extractor.v3 开始运行..

[2022-09-19 12:43:35.320524] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.322243] INFO: moduleinvoker: derived_feature_extractor.v3 运行完成[0.010226s].

[2022-09-19 12:43:35.332682] INFO: moduleinvoker: filter_stockcode.v2 开始运行..

[2022-09-19 12:43:35.340487] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.342828] INFO: moduleinvoker: filter_stockcode.v2 运行完成[0.010147s].

[2022-09-19 12:43:35.354534] INFO: moduleinvoker: filtet_st_stock.v7 开始运行..

[2022-09-19 12:43:35.362512] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.364458] INFO: moduleinvoker: filtet_st_stock.v7 运行完成[0.009926s].

[2022-09-19 12:43:35.387369] INFO: moduleinvoker: filter_delist_stocks.v3 开始运行..

[2022-09-19 12:43:35.395349] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.396876] INFO: moduleinvoker: filter_delist_stocks.v3 运行完成[0.009515s].

[2022-09-19 12:43:35.401825] INFO: moduleinvoker: standardlize.v8 开始运行..

[2022-09-19 12:43:35.408667] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.410512] INFO: moduleinvoker: standardlize.v8 运行完成[0.008681s].

[2022-09-19 12:43:35.418925] INFO: moduleinvoker: fillnan.v1 开始运行..

[2022-09-19 12:43:35.425924] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.427864] INFO: moduleinvoker: fillnan.v1 运行完成[0.008967s].

[2022-09-19 12:43:35.431790] INFO: moduleinvoker: input_features.v1 开始运行..

[2022-09-19 12:43:35.437590] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.439279] INFO: moduleinvoker: input_features.v1 运行完成[0.007494s].

[2022-09-19 12:43:35.452808] INFO: moduleinvoker: general_feature_extractor.v7 开始运行..

[2022-09-19 12:43:35.458900] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.460631] INFO: moduleinvoker: general_feature_extractor.v7 运行完成[0.007852s].

[2022-09-19 12:43:35.489102] INFO: moduleinvoker: cached.v3 开始运行..

[2022-09-19 12:43:35.496114] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.499042] INFO: moduleinvoker: cached.v3 运行完成[0.009947s].

[2022-09-19 12:43:35.510394] INFO: moduleinvoker: join.v3 开始运行..

[2022-09-19 12:43:35.517564] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.519770] INFO: moduleinvoker: join.v3 运行完成[0.009381s].

[2022-09-19 12:43:35.532929] INFO: moduleinvoker: filter.v3 开始运行..

[2022-09-19 12:43:35.541961] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.544354] INFO: moduleinvoker: filter.v3 运行完成[0.011421s].

[2022-09-19 12:43:35.553591] INFO: moduleinvoker: filter.v3 开始运行..

[2022-09-19 12:43:35.560105] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:35.562377] INFO: moduleinvoker: filter.v3 运行完成[0.00878s].

[2022-09-19 12:43:35.572681] INFO: moduleinvoker: auto_labeler_on_datasource.v1 开始运行..

[2022-09-19 12:43:35.679506] INFO: 自动标注(任意数据源): 开始标注 ..

[2022-09-19 12:43:35.844381] INFO: moduleinvoker: auto_labeler_on_datasource.v1 运行完成[0.271683s].

[2022-09-19 12:43:35.850064] INFO: moduleinvoker: standardlize.v8 开始运行..

[2022-09-19 12:43:36.469403] INFO: moduleinvoker: standardlize.v8 运行完成[0.619336s].

[2022-09-19 12:43:36.489124] INFO: moduleinvoker: join.v3 开始运行..

[2022-09-19 12:43:42.442482] INFO: join: /data, 行数=44083/749446, 耗时=5.707207s

[2022-09-19 12:43:42.498497] INFO: join: 最终行数: 44083

[2022-09-19 12:43:42.515198] INFO: moduleinvoker: join.v3 运行完成[6.026079s].

[2022-09-19 12:43:42.532161] INFO: moduleinvoker: dl_convert_to_bin.v2 开始运行..

[2022-09-19 12:43:42.998627] INFO: moduleinvoker: dl_convert_to_bin.v2 运行完成[0.466458s].

[2022-09-19 12:43:43.013588] INFO: moduleinvoker: cached.v3 开始运行..

[2022-09-19 12:43:43.442425] INFO: moduleinvoker: cached.v3 运行完成[0.42885s].

[2022-09-19 12:43:43.448870] INFO: moduleinvoker: input_features.v1 开始运行..

[2022-09-19 12:43:43.454662] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:43.456440] INFO: moduleinvoker: input_features.v1 运行完成[0.00758s].

[2022-09-19 12:43:43.467730] INFO: moduleinvoker: general_feature_extractor.v7 开始运行..

[2022-09-19 12:43:43.478368] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:43.480305] INFO: moduleinvoker: general_feature_extractor.v7 运行完成[0.012572s].

[2022-09-19 12:43:43.507040] INFO: moduleinvoker: cached.v3 开始运行..

[2022-09-19 12:43:43.514652] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:43.516997] INFO: moduleinvoker: cached.v3 运行完成[0.009982s].

[2022-09-19 12:43:43.526501] INFO: moduleinvoker: join.v3 开始运行..

[2022-09-19 12:43:43.533783] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:43.535981] INFO: moduleinvoker: join.v3 运行完成[0.009486s].

[2022-09-19 12:43:43.546529] INFO: moduleinvoker: join.v3 开始运行..

[2022-09-19 12:43:43.558485] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:43.561027] INFO: moduleinvoker: join.v3 运行完成[0.014498s].

[2022-09-19 12:43:43.583818] INFO: moduleinvoker: filter.v3 开始运行..

[2022-09-19 12:43:43.595111] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:43.597719] INFO: moduleinvoker: filter.v3 运行完成[0.013903s].

[2022-09-19 12:43:43.608660] INFO: moduleinvoker: filter.v3 开始运行..

[2022-09-19 12:43:43.616393] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:43.618923] INFO: moduleinvoker: filter.v3 运行完成[0.010269s].

[2022-09-19 12:43:43.629415] INFO: moduleinvoker: dropnan.v1 开始运行..

[2022-09-19 12:43:43.637702] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:43.639862] INFO: moduleinvoker: dropnan.v1 运行完成[0.010458s].

[2022-09-19 12:43:43.658668] INFO: moduleinvoker: dl_convert_to_bin.v2 开始运行..

[2022-09-19 12:43:43.665952] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:43.668215] INFO: moduleinvoker: dl_convert_to_bin.v2 运行完成[0.009568s].

[2022-09-19 12:43:43.692129] INFO: moduleinvoker: join.v3 开始运行..

[2022-09-19 12:43:43.698823] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:43.701314] INFO: moduleinvoker: join.v3 运行完成[0.00916s].

[2022-09-19 12:43:43.710621] INFO: moduleinvoker: dl_layer_input.v1 运行完成[0.001913s].

[2022-09-19 12:43:46.506145] INFO: moduleinvoker: dl_layer_userlayer.v1 运行完成[2.788989s].

[2022-09-19 12:43:46.523536] INFO: moduleinvoker: dl_layer_dense.v1 运行完成[0.007756s].

[2022-09-19 12:43:46.561776] INFO: moduleinvoker: cached.v3 开始运行..

[2022-09-19 12:43:46.582207] INFO: moduleinvoker: 命中缓存

[2022-09-19 12:43:46.584237] INFO: moduleinvoker: cached.v3 运行完成[0.02248s].

[2022-09-19 12:43:46.586413] INFO: moduleinvoker: dl_model_init.v1 运行完成[0.056322s].

[2022-09-19 12:43:46.592203] INFO: moduleinvoker: dl_model_train.v1 开始运行..

[2022-09-19 12:43:47.681300] INFO: dl_model_train: 准备训练，训练样本个数：35266，迭代次数：10

[2022-09-19 12:44:32.921055] INFO: dl_model_train: 训练结束，耗时：45.24s

[2022-09-19 12:44:32.965691] INFO: moduleinvoker: dl_model_train.v1 运行完成[46.373481s].

[2022-09-19 12:44:32.972617] INFO: moduleinvoker: dl_model_predict.v1 开始运行..

[2022-09-19 12:44:38.907248] INFO: moduleinvoker: dl_model_predict.v1 运行完成[5.934648s].

[2022-09-19 12:44:38.926066] INFO: moduleinvoker: cached.v3 开始运行..

[2022-09-19 12:44:40.514793] INFO: moduleinvoker: cached.v3 运行完成[1.588695s].

[2022-09-19 12:44:40.568584] INFO: moduleinvoker: backtest.v8 开始运行..

[2022-09-19 12:44:43.184187] ERROR: moduleinvoker: module name: backtest, module version: v8, trackeback: AssertionError: backtestv8: start_date param must not be None!

[2022-09-19 12:44:43.188160] ERROR: moduleinvoker: module name: trade, module version: v4, trackeback: AssertionError: backtestv8: start_date param must not be None!

[TabNet]: 32 features will be used for decision steps.

[TabNet]: 32 features will be used for decision steps.

Epoch 1/10
4/4 - 16s - loss: 0.9971 - mse: 0.9971 - val_loss: 0.9959 - val_mse: 0.9959
Epoch 2/10
4/4 - 3s - loss: 0.9958 - mse: 0.9958 - val_loss: 0.9938 - val_mse: 0.9938
Epoch 3/10
4/4 - 3s - loss: 0.9929 - mse: 0.9929 - val_loss: 0.9912 - val_mse: 0.9912
Epoch 4/10
4/4 - 3s - loss: 0.9889 - mse: 0.9889 - val_loss: 0.9868 - val_mse: 0.9868
Epoch 5/10
4/4 - 3s - loss: 0.9839 - mse: 0.9839 - val_loss: 0.9864 - val_mse: 0.9864
Epoch 6/10
4/4 - 3s - loss: 0.9817 - mse: 0.9817 - val_loss: 0.9865 - val_mse: 0.9865
Epoch 7/10
4/4 - 3s - loss: 0.9801 - mse: 0.9801 - val_loss: 0.9827 - val_mse: 0.9827
Epoch 8/10
4/4 - 3s - loss: 0.9771 - mse: 0.9771 - val_loss: 0.9794 - val_mse: 0.9794
Epoch 9/10
4/4 - 3s - loss: 0.9752 - mse: 0.9752 - val_loss: 0.9773 - val_mse: 0.9773
Epoch 10/10
4/4 - 3s - loss: 0.9734 - mse: 0.9734 - val_loss: 0.9772 - val_mse: 0.9772

[TabNet]: 32 features will be used for decision steps.
101/101 - 5s
DataSource(8fd2b8f4351b41688ae5504a33f55244T)

---------------------------------------------------------------------------
AssertionError                            Traceback (most recent call last)
<ipython-input-76-1e3a588e5c1c> in <module>
   1288 )
   1289 
-> 1290 m6 = M.trade.v4(
   1291     instruments=m2.data,
   1292     options_data=m10.data_1,

AssertionError: backtestv8: start_date param must not be None!