WARNING: These types are provided for backward compatibility only, and are not maintained anymore. They might also differ from the types supported on the backend. It is therefore strongly advised to migrate to the types found in Types for Google Cloud Bigquery v2 API.Also see the 3.0.0 Migration Guide for more information.
class google.cloud.bigquery_v2.types.DeleteModelRequest(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.Message
project_id()
Required. Project ID of the model to delete.
Type
str
dataset_id()
Required. Dataset ID of the model to delete.
Type
str
model_id()
Required. Model ID of the model to delete.
Type
str
class google.cloud.bigquery_v2.types.EncryptionConfiguration(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.Message
kms_key_name()
Optional. Describes the Cloud KMS encryption key that will be used to protect destination BigQuery table. The BigQuery Service Account associated with your project requires access to this encryption key.
Type
google.protobuf.wrappers_pb2.StringValue
class google.cloud.bigquery_v2.types.GetModelRequest(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.Message
project_id()
Required. Project ID of the requested model.
Type
str
dataset_id()
Required. Dataset ID of the requested model.
Type
str
model_id()
Required. Model ID of the requested model.
Type
str
class google.cloud.bigquery_v2.types.ListModelsRequest(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.Message
project_id()
Required. Project ID of the models to list.
Type
str
dataset_id()
Required. Dataset ID of the models to list.
Type
str
max_results()
The maximum number of results to return in a single response page. Leverage the page tokens to iterate through the entire collection.
Type
google.protobuf.wrappers_pb2.UInt32Value
page_token()
Page token, returned by a previous call to request the next page of results
Type
str
class google.cloud.bigquery_v2.types.ListModelsResponse(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.Message
models()
Models in the requested dataset. Only the following fields are populated: model_reference, model_type, creation_time, last_modified_time and labels.
Type
Sequence[google.cloud.bigquery_v2.types.Model]
next_page_token()
A token to request the next page of results.
Type
str
class google.cloud.bigquery_v2.types.Model(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.Message
etag()
Output only. A hash of this resource.
Type
str
model_reference()
Required. Unique identifier for this model.
Type
google.cloud.bigquery_v2.types.ModelReference
creation_time()
Output only. The time when this model was created, in millisecs since the epoch.
Type
int
last_modified_time()
Output only. The time when this model was last modified, in millisecs since the epoch.
Type
int
description()
Optional. A user-friendly description of this model.
Type
str
friendly_name()
Optional. A descriptive name for this model.
Type
str
labels()
The labels associated with this model. You can use these to organize and group your models. Label keys and values can be no longer than 63 characters, can only contain lowercase letters, numeric characters, underscores and dashes. International characters are allowed. Label values are optional. Label keys must start with a letter and each label in the list must have a different key.
Type
Mapping[str, str]
expiration_time()
Optional. The time when this model expires, in milliseconds since the epoch. If not present, the model will persist indefinitely. Expired models will be deleted and their storage reclaimed. The defaultTableExpirationMs property of the encapsulating dataset can be used to set a default expirationTime on newly created models.
Type
int
location()
Output only. The geographic location where the model resides. This value is inherited from the dataset.
Type
str
encryption_configuration()
Custom encryption configuration (e.g., Cloud KMS keys). This shows the encryption configuration of the model data while stored in BigQuery storage. This field can be used with PatchModel to update encryption key for an already encrypted model.
Type
google.cloud.bigquery_v2.types.EncryptionConfiguration
model_type()
Output only. Type of the model resource.
Type
google.cloud.bigquery_v2.types.Model.ModelType
training_runs()
Output only. Information for all training runs in increasing order of start_time.
Type
Sequence[google.cloud.bigquery_v2.types.Model.TrainingRun]
feature_columns()
Output only. Input feature columns that were used to train this model.
Type
Sequence[google.cloud.bigquery_v2.types.StandardSqlField]
label_columns()
Output only. Label columns that were used to train this model. The output of the model will have a predicted_ prefix to these columns.
Type
Sequence[google.cloud.bigquery_v2.types.StandardSqlField]
best_trial_id()
The best trial_id across all training runs.
Type
int
class AggregateClassificationMetrics(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageAggregate metrics for classification/classifier models. For
multi-class models, the metrics are either macro-averaged or
micro-averaged. When macro-averaged, the metrics are calculated
for each label and then an unweighted average is taken of those
values. When micro-averaged, the metric is calculated globally
by counting the total number of correctly predicted rows.
precision()
Precision is the fraction of actual positive predictions that had positive actual labels. For multiclass this is a macro-averaged metric treating each class as a binary classifier.
Type
google.protobuf.wrappers_pb2.DoubleValue
recall()
Recall is the fraction of actual positive labels that were given a positive prediction. For multiclass this is a macro-averaged metric.
Type
google.protobuf.wrappers_pb2.DoubleValue
accuracy()
Accuracy is the fraction of predictions given the correct label. For multiclass this is a micro-averaged metric.
Type
google.protobuf.wrappers_pb2.DoubleValue
threshold()
Threshold at which the metrics are computed. For binary classification models this is the positive class threshold. For multi-class classfication models this is the confidence threshold.
Type
google.protobuf.wrappers_pb2.DoubleValue
f1_score()
The F1 score is an average of recall and precision. For multiclass this is a macro-averaged metric.
Type
google.protobuf.wrappers_pb2.DoubleValue
log_loss()
Logarithmic Loss. For multiclass this is a macro-averaged metric.
Type
google.protobuf.wrappers_pb2.DoubleValue
roc_auc()
Area Under a ROC Curve. For multiclass this is a macro-averaged metric.
Type
google.protobuf.wrappers_pb2.DoubleValue
class ArimaFittingMetrics(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageARIMA model fitting metrics.
log_likelihood()
Log-likelihood.
Type
float
aic()
AIC.
Type
float
variance()
Variance.
Type
float
class ArimaForecastingMetrics(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageModel evaluation metrics for ARIMA forecasting models.
non_seasonal_order()
Non-seasonal order.
Type
Sequence[google.cloud.bigquery_v2.types.Model.ArimaOrder]
arima_fitting_metrics()
Arima model fitting metrics.
Type
Sequence[google.cloud.bigquery_v2.types.Model.ArimaFittingMetrics]
seasonal_periods()
Seasonal periods. Repeated because multiple periods are supported for one time series.
Type
Sequence[google.cloud.bigquery_v2.types.Model.SeasonalPeriod.SeasonalPeriodType]
has_drift()
Whether Arima model fitted with drift or not. It is always false when d is not 1.
Type
Sequence[bool]
time_series_id()
Id to differentiate different time series for the large-scale case.
Type
Sequence[str]
arima_single_model_forecasting_metrics()
Repeated as there can be many metric sets (one for each model) in auto-arima and the large-scale case.
Type
Sequence[google.cloud.bigquery_v2.types.Model.ArimaForecastingMetrics.ArimaSingleModelForecastingMetrics]
class ArimaSingleModelForecastingMetrics(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageModel evaluation metrics for a single ARIMA forecasting
model.
non_seasonal_order()
Non-seasonal order.
Type
google.cloud.bigquery_v2.types.Model.ArimaOrder
arima_fitting_metrics()
Arima fitting metrics.
Type
google.cloud.bigquery_v2.types.Model.ArimaFittingMetrics
has_drift()
Is arima model fitted with drift or not. It is always false when d is not 1.
Type
bool
time_series_id()
The time_series_id value for this time series. It will be one of the unique values from the time_series_id_column specified during ARIMA model training. Only present when time_series_id_column training option was used.
Type
str
time_series_ids()
The tuple of time_series_ids identifying this time series. It will be one of the unique tuples of values present in the time_series_id_columns specified during ARIMA model training. Only present when time_series_id_columns training option was used and the order of values here are same as the order of time_series_id_columns.
Type
Sequence[str]
seasonal_periods()
Seasonal periods. Repeated because multiple periods are supported for one time series.
Type
Sequence[google.cloud.bigquery_v2.types.Model.SeasonalPeriod.SeasonalPeriodType]
has_holiday_effect()
If true, holiday_effect is a part of time series decomposition result.
Type
google.protobuf.wrappers_pb2.BoolValue
has_spikes_and_dips()
If true, spikes_and_dips is a part of time series decomposition result.
Type
google.protobuf.wrappers_pb2.BoolValue
has_step_changes()
If true, step_changes is a part of time series decomposition result.
Type
google.protobuf.wrappers_pb2.BoolValue
class ArimaOrder(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageArima order, can be used for both non-seasonal and seasonal
parts.
p()
Order of the autoregressive part.
Type
int
d()
Order of the differencing part.
Type
int
q()
Order of the moving-average part.
Type
int
class BinaryClassificationMetrics(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageEvaluation metrics for binary classification/classifier
models.
aggregate_classification_metrics()
Aggregate classification metrics.
Type
google.cloud.bigquery_v2.types.Model.AggregateClassificationMetrics
binary_confusion_matrix_list()
Binary confusion matrix at multiple thresholds.
Type
Sequence[google.cloud.bigquery_v2.types.Model.BinaryClassificationMetrics.BinaryConfusionMatrix]
positive_label()
Label representing the positive class.
Type
str
negative_label()
Label representing the negative class.
Type
str
class BinaryConfusionMatrix(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageConfusion matrix for binary classification models.
positive_class_threshold()
Threshold value used when computing each of the following metric.
Type
google.protobuf.wrappers_pb2.DoubleValue
true_positives()
Number of true samples predicted as true.
Type
google.protobuf.wrappers_pb2.Int64Value
false_positives()
Number of false samples predicted as true.
Type
google.protobuf.wrappers_pb2.Int64Value
true_negatives()
Number of true samples predicted as false.
Type
google.protobuf.wrappers_pb2.Int64Value
false_negatives()
Number of false samples predicted as false.
Type
google.protobuf.wrappers_pb2.Int64Value
precision()
The fraction of actual positive predictions that had positive actual labels.
Type
google.protobuf.wrappers_pb2.DoubleValue
recall()
The fraction of actual positive labels that were given a positive prediction.
Type
google.protobuf.wrappers_pb2.DoubleValue
f1_score()
The equally weighted average of recall and precision.
Type
google.protobuf.wrappers_pb2.DoubleValue
accuracy()
The fraction of predictions given the correct label.
Type
google.protobuf.wrappers_pb2.DoubleValue
class ClusteringMetrics(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageEvaluation metrics for clustering models.
davies_bouldin_index()
Davies-Bouldin index.
Type
google.protobuf.wrappers_pb2.DoubleValue
mean_squared_distance()
Mean of squared distances between each sample to its cluster centroid.
Type
google.protobuf.wrappers_pb2.DoubleValue
clusters()
Information for all clusters.
Type
Sequence[google.cloud.bigquery_v2.types.Model.ClusteringMetrics.Cluster]
class Cluster(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageMessage containing the information about one cluster.
centroid_id()
Centroid id.
Type
int
feature_values()
Values of highly variant features for this cluster.
Type
Sequence[google.cloud.bigquery_v2.types.Model.ClusteringMetrics.Cluster.FeatureValue]
count()
Count of training data rows that were assigned to this cluster.
Type
google.protobuf.wrappers_pb2.Int64Value
class FeatureValue(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageRepresentative value of a single feature within the cluster.This message has oneof fields (mutually exclusive fields).
For each oneof, at most one member field can be set at the same time.
Setting any member of the oneof automatically clears all other
members.
feature_column()
The feature column name.
Type
str
numerical_value()
The numerical feature value. This is the
centroid value for this feature.This field is a member of oneof value.
Type
google.protobuf.wrappers_pb2.DoubleValue
categorical_value()
The categorical feature value.This field is a member of oneof value.
Type
google.cloud.bigquery_v2.types.Model.ClusteringMetrics.Cluster.FeatureValue.CategoricalValue
class CategoricalValue(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageRepresentative value of a categorical feature.
category_counts()
Counts of all categories for the categorical feature. If there are more than ten categories, we return top ten (by count) and return one more CategoryCount with category “OTHER” and count as aggregate counts of remaining categories.
Type
Sequence[google.cloud.bigquery_v2.types.Model.ClusteringMetrics.Cluster.FeatureValue.CategoricalValue.CategoryCount]
class CategoryCount(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageRepresents the count of a single category within the cluster.
category()
The name of category.
Type
str
count()
The count of training samples matching the category within the cluster.
Type
google.protobuf.wrappers_pb2.Int64Value
class DataFrequency(value)
Bases: proto.enums.EnumType of supported data frequency for time series forecasting
models.
AUTO_FREQUENCY( = )
DAILY( = )
DATA_FREQUENCY_UNSPECIFIED( = )
HOURLY( = )
MONTHLY( = )
PER_MINUTE( = )
QUARTERLY( = )
WEEKLY( = )
YEARLY( = )
class DataSplitMethod(value)
Bases: proto.enums.EnumIndicates the method to split input data into multiple
tables.
AUTO_SPLIT( = )
CUSTOM( = )
DATA_SPLIT_METHOD_UNSPECIFIED( = )
NO_SPLIT( = )
RANDOM( = )
SEQUENTIAL( = )
class DataSplitResult(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageData split result. This contains references to the training
and evaluation data tables that were used to train the model.
training_table()
Table reference of the training data after split.
Type
google.cloud.bigquery_v2.types.TableReference
evaluation_table()
Table reference of the evaluation data after split.
Type
google.cloud.bigquery_v2.types.TableReference
class DistanceType(value)
Bases: proto.enums.EnumDistance metric used to compute the distance between two
points.
COSINE( = )
DISTANCE_TYPE_UNSPECIFIED( = )
EUCLIDEAN( = )
class EvaluationMetrics(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageEvaluation metrics of a model. These are either computed on
all training data or just the eval data based on whether eval
data was used during training. These are not present for
imported models.This message has oneof fields (mutually exclusive fields).
For each oneof, at most one member field can be set at the same time.
Setting any member of the oneof automatically clears all other
members.
regression_metrics()
Populated for regression models and explicit
feedback type matrix factorization models.This field is a member of oneof metrics.
Type
google.cloud.bigquery_v2.types.Model.RegressionMetrics
binary_classification_metrics()
Populated for binary
classification/classifier models.This field is a member of oneof metrics.
Type
google.cloud.bigquery_v2.types.Model.BinaryClassificationMetrics
multi_class_classification_metrics()
Populated for multi-class
classification/classifier models.This field is a member of oneof metrics.
Type
google.cloud.bigquery_v2.types.Model.MultiClassClassificationMetrics
clustering_metrics()
Populated for clustering models.This field is a member of oneof metrics.
Type
google.cloud.bigquery_v2.types.Model.ClusteringMetrics
ranking_metrics()
Populated for implicit feedback type matrix
factorization models.This field is a member of oneof metrics.
Type
google.cloud.bigquery_v2.types.Model.RankingMetrics
arima_forecasting_metrics()
Populated for ARIMA models.This field is a member of oneof metrics.
Type
google.cloud.bigquery_v2.types.Model.ArimaForecastingMetrics
class FeedbackType(value)
Bases: proto.enums.EnumIndicates the training algorithm to use for matrix
factorization models.
EXPLICIT( = )
FEEDBACK_TYPE_UNSPECIFIED( = )
IMPLICIT( = )
class GlobalExplanation(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageGlobal explanations containing the top most important
features after training.
explanations()
A list of the top global explanations. Sorted by absolute value of attribution in descending order.
Type
Sequence[google.cloud.bigquery_v2.types.Model.GlobalExplanation.Explanation]
class_label()
Class label for this set of global explanations. Will be empty/null for binary logistic and linear regression models. Sorted alphabetically in descending order.
Type
str
class Explanation(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageExplanation for a single feature.
feature_name()
Full name of the feature. For non-numerical features, will be formatted like <column_name>.<encoded_feature_name>. Overall size of feature name will always be truncated to first 120 characters.
Type
str
attribution()
Attribution of feature.
Type
google.protobuf.wrappers_pb2.DoubleValue
class HolidayRegion(value)
Bases: proto.enums.EnumType of supported holiday regions for time series forecasting
models.
AE( = )
AR( = )
AT( = )
AU( = )
BE( = 1 )
BR( = 1 )
CA( = 1 )
CH( = 1 )
CL( = 1 )
CN( = 1 )
CO( = 1 )
CS( = 1 )
CZ( = 1 )
DE( = 1 )
DK( = 2 )
DZ( = 2 )
EC( = 2 )
EE( = 2 )
EG( = 2 )
EMEA( = )
ES( = 2 )
FI( = 2 )
FR( = 2 )
GB( = 2 )
GLOBAL( = )
GR( = 2 )
HK( = 3 )
HOLIDAY_REGION_UNSPECIFIED( = )
HU( = 3 )
ID( = 3 )
IE( = 3 )
IL( = 3 )
IN( = 3 )
IR( = 3 )
IT( = 3 )
JAPAC( = )
JP( = 3 )
KR( = 3 )
LAC( = )
LV( = 4 )
MA( = 4 )
MX( = 4 )
MY( = 4 )
NA( = )
NG( = 4 )
NL( = 4 )
NO( = 4 )
NZ( = 4 )
PE( = 4 )
PH( = 4 )
PK( = 5 )
PL( = 5 )
PT( = 5 )
RO( = 5 )
RS( = 5 )
RU( = 5 )
SA( = 5 )
SE( = 5 )
SG( = 5 )
SI( = 5 )
SK( = 6 )
TH( = 6 )
TR( = 6 )
TW( = 6 )
UA( = 6 )
US( = 6 )
VE( = 6 )
VN( = 6 )
ZA( = 6 )
class KmeansEnums(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.Message
class KmeansInitializationMethod(value)
Bases: proto.enums.EnumIndicates the method used to initialize the centroids for
KMeans clustering algorithm.
CUSTOM( = )
KMEANS_INITIALIZATION_METHOD_UNSPECIFIED( = )
KMEANS_PLUS_PLUS( = )
RANDOM( = )
class LabelsEntry(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.Message
class LearnRateStrategy(value)
Bases: proto.enums.EnumIndicates the learning rate optimization strategy to use.
CONSTANT( = )
LEARN_RATE_STRATEGY_UNSPECIFIED( = )
LINE_SEARCH( = )
class LossType(value)
Bases: proto.enums.EnumLoss metric to evaluate model training performance.
LOSS_TYPE_UNSPECIFIED( = )
MEAN_LOG_LOSS( = )
MEAN_SQUARED_LOSS( = )
class ModelType(value)
Bases: proto.enums.EnumIndicates the type of the Model.
ARIMA( = 1 )
ARIMA_PLUS( = 1 )
AUTOML_CLASSIFIER( = 1 )
AUTOML_REGRESSOR( = 1 )
BOOSTED_TREE_CLASSIFIER( = 1 )
BOOSTED_TREE_REGRESSOR( = )
DNN_CLASSIFIER( = )
DNN_REGRESSOR( = )
KMEANS( = )
LINEAR_REGRESSION( = )
LOGISTIC_REGRESSION( = )
MATRIX_FACTORIZATION( = )
MODEL_TYPE_UNSPECIFIED( = )
TENSORFLOW( = )
class MultiClassClassificationMetrics(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageEvaluation metrics for multi-class classification/classifier
models.
aggregate_classification_metrics()
Aggregate classification metrics.
Type
google.cloud.bigquery_v2.types.Model.AggregateClassificationMetrics
confusion_matrix_list()
Confusion matrix at different thresholds.
Type
Sequence[google.cloud.bigquery_v2.types.Model.MultiClassClassificationMetrics.ConfusionMatrix]
class ConfusionMatrix(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageConfusion matrix for multi-class classification models.
confidence_threshold()
Confidence threshold used when computing the entries of the confusion matrix.
Type
google.protobuf.wrappers_pb2.DoubleValue
rows()
One row per actual label.
Type
Sequence[google.cloud.bigquery_v2.types.Model.MultiClassClassificationMetrics.ConfusionMatrix.Row]
class Entry(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageA single entry in the confusion matrix.
predicted_label()
The predicted label. For confidence_threshold > 0, we will also add an entry indicating the number of items under the confidence threshold.
Type
str
item_count()
Number of items being predicted as this label.
Type
google.protobuf.wrappers_pb2.Int64Value
class Row(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageA single row in the confusion matrix.
actual_label()
The original label of this row.
Type
str
entries()
Info describing predicted label distribution.
Type
Sequence[google.cloud.bigquery_v2.types.Model.MultiClassClassificationMetrics.ConfusionMatrix.Entry]
class OptimizationStrategy(value)
Bases: proto.enums.EnumIndicates the optimization strategy used for training.
BATCH_GRADIENT_DESCENT( = )
NORMAL_EQUATION( = )
OPTIMIZATION_STRATEGY_UNSPECIFIED( = )
class RankingMetrics(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageEvaluation metrics used by weighted-ALS models specified by
feedback_type=implicit.
mean_average_precision()
Calculates a precision per user for all the items by ranking them and then averages all the precisions across all the users.
Type
google.protobuf.wrappers_pb2.DoubleValue
mean_squared_error()
Similar to the mean squared error computed in regression and explicit recommendation models except instead of computing the rating directly, the output from evaluate is computed against a preference which is 1 or 0 depending on if the rating exists or not.
Type
google.protobuf.wrappers_pb2.DoubleValue
normalized_discounted_cumulative_gain()
A metric to determine the goodness of a ranking calculated from the predicted confidence by comparing it to an ideal rank measured by the original ratings.
Type
google.protobuf.wrappers_pb2.DoubleValue
average_rank()
Determines the goodness of a ranking by computing the percentile rank from the predicted confidence and dividing it by the original rank.
Type
google.protobuf.wrappers_pb2.DoubleValue
class RegressionMetrics(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageEvaluation metrics for regression and explicit feedback type
matrix factorization models.
mean_absolute_error()
Mean absolute error.
Type
google.protobuf.wrappers_pb2.DoubleValue
mean_squared_error()
Mean squared error.
Type
google.protobuf.wrappers_pb2.DoubleValue
mean_squared_log_error()
Mean squared log error.
Type
google.protobuf.wrappers_pb2.DoubleValue
median_absolute_error()
Median absolute error.
Type
google.protobuf.wrappers_pb2.DoubleValue
r_squared()
R^2 score. This corresponds to r2_score in ML.EVALUATE.
Type
google.protobuf.wrappers_pb2.DoubleValue
class SeasonalPeriod(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.Message
class SeasonalPeriodType(value)
Bases: proto.enums.Enum
DAILY( = )
MONTHLY( = )
NO_SEASONALITY( = )
QUARTERLY( = )
SEASONAL_PERIOD_TYPE_UNSPECIFIED( = )
WEEKLY( = )
YEARLY( = )
class TrainingRun(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageInformation about a single training query run for the model.
training_options()
Options that were used for this training run, includes user specified and default options that were used.
Type
google.cloud.bigquery_v2.types.Model.TrainingRun.TrainingOptions
start_time()
The start time of this training run.
Type
google.protobuf.timestamp_pb2.Timestamp
results()
Output of each iteration run, results.size() <= max_iterations.
Type
Sequence[google.cloud.bigquery_v2.types.Model.TrainingRun.IterationResult]
evaluation_metrics()
The evaluation metrics over training/eval data that were computed at the end of training.
Type
google.cloud.bigquery_v2.types.Model.EvaluationMetrics
data_split_result()
Data split result of the training run. Only set when the input data is actually split.
Type
google.cloud.bigquery_v2.types.Model.DataSplitResult
global_explanations()
Global explanations for important features of the model. For multi-class models, there is one entry for each label class. For other models, there is only one entry in the list.
Type
Sequence[google.cloud.bigquery_v2.types.Model.GlobalExplanation]
class IterationResult(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageInformation about a single iteration of the training run.
index()
Index of the iteration, 0 based.
Type
google.protobuf.wrappers_pb2.Int32Value
duration_ms()
Time taken to run the iteration in milliseconds.
Type
google.protobuf.wrappers_pb2.Int64Value
training_loss()
Loss computed on the training data at the end of iteration.
Type
google.protobuf.wrappers_pb2.DoubleValue
eval_loss()
Loss computed on the eval data at the end of iteration.
Type
google.protobuf.wrappers_pb2.DoubleValue
learn_rate()
Learn rate used for this iteration.
Type
float
cluster_infos()
Information about top clusters for clustering models.
Type
Sequence[google.cloud.bigquery_v2.types.Model.TrainingRun.IterationResult.ClusterInfo]
arima_result()
Type
google.cloud.bigquery_v2.types.Model.TrainingRun.IterationResult.ArimaResult
class ArimaResult(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.Message(Auto-)arima fitting result. Wrap everything in ArimaResult
for easier refactoring if we want to use model-specific
iteration results.
arima_model_info()
This message is repeated because there are multiple arima models fitted in auto-arima. For non-auto-arima model, its size is one.
Type
Sequence[google.cloud.bigquery_v2.types.Model.TrainingRun.IterationResult.ArimaResult.ArimaModelInfo]
seasonal_periods()
Seasonal periods. Repeated because multiple periods are supported for one time series.
Type
Sequence[google.cloud.bigquery_v2.types.Model.SeasonalPeriod.SeasonalPeriodType]
class ArimaCoefficients(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageArima coefficients.
auto_regressive_coefficients()
Auto-regressive coefficients, an array of double.
Type
Sequence[float]
moving_average_coefficients()
Moving-average coefficients, an array of double.
Type
Sequence[float]
intercept_coefficient()
Intercept coefficient, just a double not an array.
Type
float
class ArimaModelInfo(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageArima model information.
non_seasonal_order()
Non-seasonal order.
Type
google.cloud.bigquery_v2.types.Model.ArimaOrder
arima_coefficients()
Arima coefficients.
Type
google.cloud.bigquery_v2.types.Model.TrainingRun.IterationResult.ArimaResult.ArimaCoefficients
arima_fitting_metrics()
Arima fitting metrics.
Type
google.cloud.bigquery_v2.types.Model.ArimaFittingMetrics
has_drift()
Whether Arima model fitted with drift or not. It is always false when d is not 1.
Type
bool
time_series_id()
The time_series_id value for this time series. It will be one of the unique values from the time_series_id_column specified during ARIMA model training. Only present when time_series_id_column training option was used.
Type
str
time_series_ids()
The tuple of time_series_ids identifying this time series. It will be one of the unique tuples of values present in the time_series_id_columns specified during ARIMA model training. Only present when time_series_id_columns training option was used and the order of values here are same as the order of time_series_id_columns.
Type
Sequence[str]
seasonal_periods()
Seasonal periods. Repeated because multiple periods are supported for one time series.
Type
Sequence[google.cloud.bigquery_v2.types.Model.SeasonalPeriod.SeasonalPeriodType]
has_holiday_effect()
If true, holiday_effect is a part of time series decomposition result.
Type
google.protobuf.wrappers_pb2.BoolValue
has_spikes_and_dips()
If true, spikes_and_dips is a part of time series decomposition result.
Type
google.protobuf.wrappers_pb2.BoolValue
has_step_changes()
If true, step_changes is a part of time series decomposition result.
Type
google.protobuf.wrappers_pb2.BoolValue
class ClusterInfo(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageInformation about a single cluster for clustering model.
centroid_id()
Centroid id.
Type
int
cluster_radius()
Cluster radius, the average distance from centroid to each point assigned to the cluster.
Type
google.protobuf.wrappers_pb2.DoubleValue
cluster_size()
Cluster size, the total number of points assigned to the cluster.
Type
google.protobuf.wrappers_pb2.Int64Value
class TrainingOptions(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageOptions used in model training.
max_iterations()
The maximum number of iterations in training. Used only for iterative training algorithms.
Type
int
loss_type()
Type of loss function used during training run.
Type
google.cloud.bigquery_v2.types.Model.LossType
learn_rate()
Learning rate in training. Used only for iterative training algorithms.
Type
float
l1_regularization()
L1 regularization coefficient.
Type
google.protobuf.wrappers_pb2.DoubleValue
l2_regularization()
L2 regularization coefficient.
Type
google.protobuf.wrappers_pb2.DoubleValue
min_relative_progress()
When early_stop is true, stops training when accuracy improvement is less than ‘min_relative_progress’. Used only for iterative training algorithms.
Type
google.protobuf.wrappers_pb2.DoubleValue
warm_start()
Whether to train a model from the last checkpoint.
Type
google.protobuf.wrappers_pb2.BoolValue
early_stop()
Whether to stop early when the loss doesn’t improve significantly any more (compared to min_relative_progress). Used only for iterative training algorithms.
Type
google.protobuf.wrappers_pb2.BoolValue
input_label_columns()
Name of input label columns in training data.
Type
Sequence[str]
data_split_method()
The data split type for training and evaluation, e.g. RANDOM.
Type
google.cloud.bigquery_v2.types.Model.DataSplitMethod
data_split_eval_fraction()
The fraction of evaluation data over the whole input data. The rest of data will be used as training data. The format should be double. Accurate to two decimal places. Default value is 0.2.
Type
float
data_split_column()
The column to split data with. This column won’t be used as a feature.
- When data_split_method is CUSTOM, the corresponding column should be boolean. The rows with true value tag are eval data, and the false are training data.
- When data_split_method is SEQ, the first DATA_SPLIT_EVAL_FRACTION rows (from smallest to largest) in the corresponding column are used as training data, and the rest are eval data. It respects the order in Orderable data types: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#data-type-properties
Type
str
learn_rate_strategy()
The strategy to determine learn rate for the current iteration.
Type
google.cloud.bigquery_v2.types.Model.LearnRateStrategy
initial_learn_rate()
Specifies the initial learning rate for the line search learn rate strategy.
Type
float
label_class_weights()
Weights associated with each label class, for rebalancing the training data. Only applicable for classification models.
Type
Mapping[str, float]
user_column()
User column specified for matrix factorization models.
Type
str
item_column()
Item column specified for matrix factorization models.
Type
str
distance_type()
Distance type for clustering models.
Type
google.cloud.bigquery_v2.types.Model.DistanceType
num_clusters()
Number of clusters for clustering models.
Type
int
model_uri()
Google Cloud Storage URI from which the model was imported. Only applicable for imported models.
Type
str
optimization_strategy()
Optimization strategy for training linear regression models.
Type
google.cloud.bigquery_v2.types.Model.OptimizationStrategy
hidden_units()
Hidden units for dnn models.
Type
Sequence[int]
batch_size()
Batch size for dnn models.
Type
int
dropout()
Dropout probability for dnn models.
Type
google.protobuf.wrappers_pb2.DoubleValue
max_tree_depth()
Maximum depth of a tree for boosted tree models.
Type
int
subsample()
Subsample fraction of the training data to grow tree to prevent overfitting for boosted tree models.
Type
float
min_split_loss()
Minimum split loss for boosted tree models.
Type
google.protobuf.wrappers_pb2.DoubleValue
num_factors()
Num factors specified for matrix factorization models.
Type
int
feedback_type()
Feedback type that specifies which algorithm to run for matrix factorization.
Type
google.cloud.bigquery_v2.types.Model.FeedbackType
wals_alpha()
Hyperparameter for matrix factoration when implicit feedback type is specified.
Type
google.protobuf.wrappers_pb2.DoubleValue
kmeans_initialization_method()
The method used to initialize the centroids for kmeans algorithm.
Type
google.cloud.bigquery_v2.types.Model.KmeansEnums.KmeansInitializationMethod
kmeans_initialization_column()
The column used to provide the initial centroids for kmeans algorithm when kmeans_initialization_method is CUSTOM.
Type
str
time_series_timestamp_column()
Column to be designated as time series timestamp for ARIMA model.
Type
str
time_series_data_column()
Column to be designated as time series data for ARIMA model.
Type
str
auto_arima()
Whether to enable auto ARIMA or not.
Type
bool
non_seasonal_order()
A specification of the non-seasonal part of the ARIMA model: the three components (p, d, q) are the AR order, the degree of differencing, and the MA order.
Type
google.cloud.bigquery_v2.types.Model.ArimaOrder
data_frequency()
The data frequency of a time series.
Type
google.cloud.bigquery_v2.types.Model.DataFrequency
include_drift()
Include drift when fitting an ARIMA model.
Type
bool
holiday_region()
The geographical region based on which the holidays are considered in time series modeling. If a valid value is specified, then holiday effects modeling is enabled.
Type
google.cloud.bigquery_v2.types.Model.HolidayRegion
time_series_id_column()
The time series id column that was used during ARIMA model training.
Type
str
time_series_id_columns()
The time series id columns that were used during ARIMA model training.
Type
Sequence[str]
horizon()
The number of periods ahead that need to be forecasted.
Type
int
preserve_input_structs()
Whether to preserve the input structs in output feature names. Suppose there is a struct A with field b. When false (default), the output feature name is A_b. When true, the output feature name is A.b.
Type
bool
auto_arima_max_order()
The max value of non-seasonal p and q.
Type
int
decompose_time_series()
If true, perform decompose time series and save the results.
Type
google.protobuf.wrappers_pb2.BoolValue
clean_spikes_and_dips()
If true, clean spikes and dips in the input time series.
Type
google.protobuf.wrappers_pb2.BoolValue
adjust_step_changes()
If true, detect step changes and make data adjustment in the input time series.
Type
google.protobuf.wrappers_pb2.BoolValue
class LabelClassWeightsEntry(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.Message
class google.cloud.bigquery_v2.types.ModelReference(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageId path of a model.
project_id()
Required. The ID of the project containing this model.
Type
str
dataset_id()
Required. The ID of the dataset containing this model.
Type
str
model_id()
Required. The ID of the model. The ID must contain only letters (a-z, A-Z), numbers (0-9), or underscores (_). The maximum length is 1,024 characters.
Type
str
class google.cloud.bigquery_v2.types.PatchModelRequest(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.Message
project_id()
Required. Project ID of the model to patch.
Type
str
dataset_id()
Required. Dataset ID of the model to patch.
Type
str
model_id()
Required. Model ID of the model to patch.
Type
str
model()
Required. Patched model. Follows RFC5789 patch semantics. Missing fields are not updated. To clear a field, explicitly set to default value.
Type
google.cloud.bigquery_v2.types.Model
class google.cloud.bigquery_v2.types.StandardSqlDataType(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageThe type of a variable, e.g., a function argument. Examples: INT64:
{type_kind=”INT64”} ARRAY: {type_kind=”ARRAY”,
array_element_type=”STRING”} STRUCT<x STRING, y ARRAY>:
{type_kind=”STRUCT”, struct_type={fields=[ {name=”x”,
type={type_kind=”STRING”}}, {name=”y”, type={type_kind=”ARRAY”,
array_element_type=”DATE”}} ]}}This message has oneof fields (mutually exclusive fields).
For each oneof, at most one member field can be set at the same time.
Setting any member of the oneof automatically clears all other
members.
type_kind()
Required. The top level type of this field. Can be any standard SQL data type (e.g., “INT64”, “DATE”, “ARRAY”).
Type
google.cloud.bigquery_v2.types.StandardSqlDataType.TypeKind
array_element_type()
The type of the array’s elements, if type_kind = “ARRAY”.This field is a member of oneof sub_type.
Type
google.cloud.bigquery_v2.types.StandardSqlDataType
struct_type()
The fields of this struct, in order, if type_kind =
“STRUCT”.This field is a member of oneof sub_type.
Type
google.cloud.bigquery_v2.types.StandardSqlStructType
class TypeKind(value)
Bases: proto.enums.Enum
ARRAY( = 1 )
BIGNUMERIC( = 2 )
BOOL( = )
BYTES( = )
DATE( = 1 )
DATETIME( = 2 )
FLOAT64( = )
GEOGRAPHY( = 2 )
INT64( = )
INTERVAL( = 2 )
JSON( = 2 )
NUMERIC( = 2 )
STRING( = )
STRUCT( = 1 )
TIME( = 2 )
TIMESTAMP( = 1 )
TYPE_KIND_UNSPECIFIED( = )
class google.cloud.bigquery_v2.types.StandardSqlField(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageA field or a column.
name()
Optional. The name of this field. Can be absent for struct fields.
Type
str
type()
Optional. The type of this parameter. Absent if not explicitly specified (e.g., CREATE FUNCTION statement can omit the return type; in this case the output parameter does not have this “type” field).
Type
google.cloud.bigquery_v2.types.StandardSqlDataType
class google.cloud.bigquery_v2.types.StandardSqlStructType(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.Message
fields()
Type
Sequence[google.cloud.bigquery_v2.types.StandardSqlField]
class google.cloud.bigquery_v2.types.StandardSqlTableType(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.MessageA table type
columns()
The columns in this table type
Type
Sequence[google.cloud.bigquery_v2.types.StandardSqlField]
class google.cloud.bigquery_v2.types.TableReference(mapping=None, *, ignore_unknown_fields=False, **kwargs)
Bases: proto.message.Message
project_id()
Required. The ID of the project containing this table.
Type
str
dataset_id()
Required. The ID of the dataset containing this table.
Type
str
table_id()
Required. The ID of the table. The ID must contain only
letters (a-z, A-Z), numbers (0-9), or underscores (_). The
maximum length is 1,024 characters. Certain operations allow
suffixing of the table ID with a partition decorator, such
as sample_table$20190123.
Type
str
project_id_alternative()
The alternative field that will be used when ESF is not able to translate the received data to the project_id field.
Type
Sequence[str]
dataset_id_alternative()
The alternative field that will be used when ESF is not able to translate the received data to the project_id field.
Type
Sequence[str]
table_id_alternative()
The alternative field that will be used when ESF is not able to translate the received data to the project_id field.
Type
Sequence[str]