class*, n_components=2, perplexity=30.0, early_exaggeration=12.0, learning_rate='auto', n_iter=1000, n_iter_without_progress=300, min_grad_norm=1e-07, metric='euclidean', metric_params=None, init='pca', verbose=0, random_state=None, method='barnes_hut', angle=0.5, n_jobs=None, input_cols: Optional[Union[str, Iterable[str]]] = None, output_cols: Optional[Union[str, Iterable[str]]] = None, label_cols: Optional[Union[str, Iterable[str]]] = None, passthrough_cols: Optional[Union[str, Iterable[str]]] = None, drop_input_cols: Optional[bool] = False, sample_weight_col: Optional[str] = None)

Bases: BaseTransformer

T-distributed Stochastic Neighbor Embedding For more details on this class, see sklearn.manifold.TSNE

  • input_cols (Optional[Union[str, List[str]]]) – A string or list of strings representing column names that contain features. If this parameter is not specified, all columns in the input DataFrame except the columns specified by label_cols, sample_weight_col, and passthrough_cols parameters are considered input columns. Input columns can also be set after initialization with the set_input_cols method.

  • label_cols (Optional[Union[str, List[str]]]) – A string or list of strings representing column names that contain labels. Label columns must be specified with this parameter during initialization or with the set_label_cols method before fitting.

  • output_cols (Optional[Union[str, List[str]]]) – A string or list of strings representing column names that will store the output of predict and transform operations. The length of output_cols must match the expected number of output columns from the specific predictor or transformer class used. If you omit this parameter, output column names are derived by adding an OUTPUT_ prefix to the label column names for supervised estimators, or OUTPUT_<IDX>for unsupervised estimators. These inferred output column names work for predictors, but output_cols must be set explicitly for transformers. In general, explicitly specifying output column names is clearer, especially if you don’t specify the input column names. To transform in place, pass the same names for input_cols and output_cols. be set explicitly for transformers. Output columns can also be set after initialization with the set_output_cols method.

  • sample_weight_col (Optional[str]) – A string representing the column name containing the sample weights. This argument is only required when working with weighted datasets. Sample weight column can also be set after initialization with the set_sample_weight_col method.

  • passthrough_cols (Optional[Union[str, List[str]]]) – A string or a list of strings indicating column names to be excluded from any operations (such as train, transform, or inference). These specified column(s) will remain untouched throughout the process. This option is helpful in scenarios requiring automatic input_cols inference, but need to avoid using specific columns, like index columns, during training or inference. Passthrough columns can also be set after initialization with the set_passthrough_cols method.

  • drop_input_cols (Optional[bool], default=False) – If set, the response of predict(), transform() methods will not contain input columns.

  • n_components (int, default=2) – Dimension of the embedded space.

  • perplexity (float, default=30.0) – The perplexity is related to the number of nearest neighbors that is used in other manifold learning algorithms. Larger datasets usually require a larger perplexity. Consider selecting a value between 5 and 50. Different values can result in significantly different results. The perplexity must be less than the number of samples.

  • early_exaggeration (float, default=12.0) – Controls how tight natural clusters in the original space are in the embedded space and how much space will be between them. For larger values, the space between natural clusters will be larger in the embedded space. Again, the choice of this parameter is not very critical. If the cost function increases during initial optimization, the early exaggeration factor or the learning rate might be too high.

  • learning_rate (float or "auto", default="auto") – The learning rate for t-SNE is usually in the range [10.0, 1000.0]. If the learning rate is too high, the data may look like a ‘ball’ with any point approximately equidistant from its nearest neighbours. If the learning rate is too low, most points may look compressed in a dense cloud with few outliers. If the cost function gets stuck in a bad local minimum increasing the learning rate may help. Note that many other t-SNE implementations (bhtsne, FIt-SNE, openTSNE, etc.) use a definition of learning_rate that is 4 times smaller than ours. So our learning_rate=200 corresponds to learning_rate=800 in those other implementations. The ‘auto’ option sets the learning_rate to max(N / early_exaggeration / 4, 50) where N is the sample size, following [4] and [5].

  • n_iter (int, default=1000) – Maximum number of iterations for the optimization. Should be at least 250.

  • n_iter_without_progress (int, default=300) – Maximum number of iterations without progress before we abort the optimization, used after 250 initial iterations with early exaggeration. Note that progress is only checked every 50 iterations so this value is rounded to the next multiple of 50.

  • min_grad_norm (float, default=1e-7) – If the gradient norm is below this threshold, the optimization will be stopped.

  • metric (str or callable, default='euclidean') – The metric to use when calculating distance between instances in a feature array. If metric is a string, it must be one of the options allowed by scipy.spatial.distance.pdist for its metric parameter, or a metric listed in pairwise.PAIRWISE_DISTANCE_FUNCTIONS. If metric is “precomputed”, X is assumed to be a distance matrix. Alternatively, if metric is a callable function, it is called on each pair of instances (rows) and the resulting value recorded. The callable should take two arrays from X as input and return a value indicating the distance between them. The default is “euclidean” which is interpreted as squared euclidean distance.

  • metric_params (dict, default=None) – Additional keyword arguments for the metric function.

  • init ({"random", "pca"} or ndarray of shape (n_samples, n_components), default="pca") – Initialization of embedding. PCA initialization cannot be used with precomputed distances and is usually more globally stable than random initialization.

  • verbose (int, default=0) – Verbosity level.

  • random_state (int, RandomState instance or None, default=None) – Determines the random number generator. Pass an int for reproducible results across multiple function calls. Note that different initializations might result in different local minima of the cost function. See Glossary.

  • method ({'barnes_hut', 'exact'}, default='barnes_hut') – By default the gradient calculation algorithm uses Barnes-Hut approximation running in O(NlogN) time. method=’exact’ will run on the slower, but exact, algorithm in O(N^2) time. The exact algorithm should be used when nearest-neighbor errors need to be better than 3%. However, the exact method cannot scale to millions of examples.

  • angle (float, default=0.5) – Only used if method=’barnes_hut’ This is the trade-off between speed and accuracy for Barnes-Hut T-SNE. ‘angle’ is the angular size (referred to as theta in [3]) of a distant node as measured from a point. If this size is below ‘angle’ then it is used as a summary node of all points contained within it. This method is not very sensitive to changes in this parameter in the range of 0.2 - 0.8. Angle less than 0.2 has quickly increasing computation time and angle greater 0.8 has quickly increasing error.

  • n_jobs (int, default=None) – The number of parallel jobs to run for neighbors search. This parameter has no impact when metric="precomputed" or (metric="euclidean" and method="exact"). None means 1 unless in a joblib.parallel_backend context. -1 means using all processors. See Glossary for more details.

Base class for all transformers.


fit(dataset: Union[DataFrame, DataFrame]) BaseEstimator

Runs universal logics for all fit implementations.

fit_transform(dataset: Union[DataFrame, DataFrame], output_cols_prefix: str = 'fit_transform_') Union[DataFrame, DataFrame]

Fit X into an embedded space and return that transformed output For more details on this function, see sklearn.manifold.TSNE.fit_transform


TypeError – Supported dataset types: snowpark.DataFrame, pandas.DataFrame.


dataset – Union[snowflake.snowpark.DataFrame, pandas.DataFrame] Snowpark or Pandas DataFrame.

output_cols_prefix: Prefix for the response columns :returns: Transformed dataset.

get_input_cols() List[str]

Input columns getter.


Input columns.

get_label_cols() List[str]

Label column getter.


Label column(s).

get_output_cols() List[str]

Output columns getter.


Output columns.

get_params(deep: bool = True) Dict[str, Any]

Get parameters for this transformer.


deep – If True, will return the parameters for this transformer and contained subobjects that are transformers.


Parameter names mapped to their values.

get_passthrough_cols() List[str]

Passthrough columns getter.


Passthrough column(s).

get_sample_weight_col() Optional[str]

Sample weight column getter.


Sample weight column.

get_sklearn_args(default_sklearn_obj: Optional[object] = None, sklearn_initial_keywords: Optional[Union[str, Iterable[str]]] = None, sklearn_unused_keywords: Optional[Union[str, Iterable[str]]] = None, snowml_only_keywords: Optional[Union[str, Iterable[str]]] = None, sklearn_added_keyword_to_version_dict: Optional[Dict[str, str]] = None, sklearn_added_kwarg_value_to_version_dict: Optional[Dict[str, Dict[str, str]]] = None, sklearn_deprecated_keyword_to_version_dict: Optional[Dict[str, str]] = None, sklearn_removed_keyword_to_version_dict: Optional[Dict[str, str]] = None) Dict[str, Any]

Get sklearn keyword arguments.

This method enables modifying object parameters for special cases.

  • default_sklearn_obj – Sklearn object used to get default parameter values. Necessary when sklearn_added_keyword_to_version_dict is provided.

  • sklearn_initial_keywords – Initial keywords in sklearn.

  • sklearn_unused_keywords – Sklearn keywords that are unused in snowml.

  • snowml_only_keywords – snowml only keywords not present in sklearn.

  • sklearn_added_keyword_to_version_dict – Added keywords mapped to the sklearn versions in which they were added.

  • sklearn_added_kwarg_value_to_version_dict – Added keyword argument values mapped to the sklearn versions in which they were added.

  • sklearn_deprecated_keyword_to_version_dict – Deprecated keywords mapped to the sklearn versions in which they were deprecated.

  • sklearn_removed_keyword_to_version_dict – Removed keywords mapped to the sklearn versions in which they were removed.


Sklearn parameter names mapped to their values.

score_samples(dataset: Union[DataFrame, DataFrame], output_cols_prefix: str = 'score_samples_') Union[DataFrame, DataFrame]

Method not supported for this class.


TypeError – Supported dataset types: snowpark.DataFrame, pandas.DataFrame.

  • dataset – Union[snowflake.snowpark.DataFrame, pandas.DataFrame] Snowpark or Pandas DataFrame.

  • output_cols_prefix – Prefix for the response columns


Output dataset with probability of the sample for each class in the model.

set_drop_input_cols(drop_input_cols: Optional[bool] = False) None
set_input_cols(input_cols: Optional[Union[str, Iterable[str]]]) TSNE

Input columns setter.


input_cols – A single input column or multiple input columns.



set_label_cols(label_cols: Optional[Union[str, Iterable[str]]]) Base

Label column setter.


label_cols – A single label column or multiple label columns if multi task learning.



set_output_cols(output_cols: Optional[Union[str, Iterable[str]]]) Base

Output columns setter.


output_cols – A single output column or multiple output columns.



set_params(**params: Dict[str, Any]) None

Set the parameters of this transformer.

The method works on simple transformers as well as on nested objects. The latter have parameters of the form <component>__<parameter> so that it’s possible to update each component of a nested object.


**params – Transformer parameter names mapped to their values.


SnowflakeMLException – Invalid parameter keys.

set_passthrough_cols(passthrough_cols: Optional[Union[str, Iterable[str]]]) Base

Passthrough columns setter.


passthrough_cols – Column(s) that should not be used or modified by the estimator/transformer. Estimator/Transformer just passthrough these columns without any modifications.



set_sample_weight_col(sample_weight_col: Optional[str]) Base

Sample weight column setter.


sample_weight_col – A single column that represents sample weight.



to_sklearn() Any

Get sklearn.manifold.TSNE object.



Returns model signature of current class.


SnowflakeMLException – If estimator is not fitted, then model signature cannot be inferred


Dict with each method and its input output signature