Identify clusters using the distance between provided embeddings.
Eqivalent to cluster_dataset, but instead of a dataset expects a column of embeddings as input. The input may e.g.
be word2vec embeddings from an embed_text step, or whole dataset embeddings from an embed_dataset step.
Optionally reduces the dimensionality of the embeddings (by default using UMAP).
This may help with making the data denser (counteracting the “curse-of-dimensionality”),
and thus making it potentially easier to identify clusters.
The clustering algorithm used by default is (HDBSCAN), which
produces a column of positive cluster IDs, or -1 if a data point is considered noise (not belonging to any cluster).
The following are the inputs expected by the step and the outputs it produces. These are generally
columns (ds.first_name), datasets (ds or ds[["first_name", "last_name"]]) or models (referenced
by name e.g. "churn-clf").
One or two columns containing the clustering results. If one column name is provided, the single output
will contain the cluster labels. If two column names are provided, the second column will contain the
the probability that a data points belongs to the assigned cluster.
The following parameters can be used to configure the behaviour of the step by including them in
a json object as the last “input” to the step, i.e. step(..., {"param": "value", ...}) -> (output).
The larger the value, the more conservative the clustering.
More points will be declared as noise, and clusters will be restricted to progressively more dense areas.
Toggle encoding of feature columns.
When enabled, Graphext will auto-convert any column types to the numeric type before
(optionally) reducing the data’s dimensionality. How this conversion is done can be
configured using the feature_encoder option below.
If disabled, the dimensionality reduction algorithm applied in this step will
assume that input data is already numerical and doesn’t contain any missing
values.
Configures encoding of feature columns.
By default (null), Graphext chooses automatically how to convert any column types the model
may not understand natively to a numeric type.
A configuration object can be passed instead to overwrite specific parameter values with respect
to their default values.
Category encoder.
May contain either a single configuration for all categorical variables, or two different configurations
for low- and high-cardinality variables. For further details pick one of the two options below.
Maximum number of unique categories to encode.
Only the N-1 most common categories will be encoded, and the rest will be grouped into a single
“Others” category.
Multilabel encoder.
Configures encoding of multivalued categorical features (variable length lists of categories,
or the semantic type list[category] for short). May contain either a single configuration for
all multilabel variables, or two different configurations for low- and high-cardinality variables.
For further details pick one of the two options below.
Maximum number of categories/labels to encode.
If a number is provided, the result of the encoding will be reduced to these many dimensions (columns)
using scikit-learn’s truncated SVD.
When applied together with (after a) Tf-Idf encoding, this performs a kind of
latent semantic analysis.
A list of cyclical time features to extract.
“Cycles” are numerical transformations of features that should be represented on a circle. E.g. months,
ranging from 1 to 12, should be arranged such that 12 and 1 are next to each other, rather than on
opposite ends of a linear scale. We represent such cyclical time features on a circle by creating two
columns for each original feature: the sin and cos of the numerical feature after appropriate scaling.
Embedding/vector encoder.
Configures encoding of multivalued numerical features (variable length lists of numbers, i.e. vectors, or the semantic type list[number] for short).
Text encoder.
Configures encoding of text (natural language) features. Currently only allows
Tf-Idf embeddings to represent texts. If you wish
to use other embeddings, e.g. semantic, Word2Vec etc., transform your text column first using
another step, and then use that result instead of the original texts.
Texts are excluded by default from the overall encoding of the dataset. See parameter
include_text_features below to active it.
Parameters to be passed to the text encoder (Tf-Idf parameters only for now).
See scikit-learn’s documentation
for detailed parameters and their explanation.
How many output features to generate.
The resulting Tf-Idf vectors will be reduced to these many dimensions (columns) using scikit-learn’s
truncated SVD.
This performs a kind of latent semantic analysis.
By default we will reduce to 200 components.
Whether to include or ignore text columns during the processing of input data.
Enabling this will convert texts to their Tf-Idf representation. Each text will be
converted to an N-dimensional vector in which each component measures the relative
“over-representation” of a specific word (or n-gram) relative to its overall
frequency in the whole dataset. This is disabled by default because it will
often be better to convert texts explicitly using a previous step, such as
embed_text or embed_text_with_model.
Weights used to multiply the normalized columns/features after vectorization.
Should be a dictionary/object of {"column_name": weight, ...} items. Will be scaled using the
parameters weights_max, and weights_exp before being applied. So only the relative weight of
the columns is important here, not their absolute values.
Weights used to multiply the normalized columns/features after vectorization.
Should be a dictionary/object of "type": weight" items. Will be scaled using the parameters
weights_max, and weights_exp before being applied. So only the relative weight of the columns
is important here, not their absolute values.
Weight exponent.
Weights will be raised to this power before(!) scaling to weights_max. This allows for a non-linear
mapping from input weights to those used eventually to multiply the normalized columns.
Minimum distance between reduced data points.
Controls how tightly UMAP is allowed to pack points together in the reduced space. Smaller values will
lead to points more tightly packed together (potentially useful if result is used to cluster the points).
Larger values will distribute points with more space between them (which may be desirable for visualization,
or to focus more on the global structure of the date).
Number of training iterations used in optimizing the embedding.
Larger values result in more accurate embeddings. If null is specified a value will be
selected based on the size of the input dataset (200 for large datasets, 500 for small).
How to initialize the low dimensional embedding.
When “spectral”, uses a (relatively expensive) spectral embedding. “pca” uses the first n_components
from a principal component analysis. “tswspectral” is a cheaper alternative to “spectral”. When “random”,
assigns initial embedding positions at random. This uses the least amount of memory and time but may make UMAP
slower to converge on the optimal embedding. “auto” selects between “spectral” and “random” automatically
depending on the size of the dataset.
Avoid excessive memory use.
For some datasets nearest neighbor computations can consume a lot of memory. If you find
the step is failing due to memory constraints, consider setting this option to true.
This approach is more computationally expensive, but avoids excessive memory use. Setting
it to “auto”, will enable this mode automatically depending on the size of the dataset.
Weighting factor between features and target.
A value of 0.0 weights entirely on data, and a value of 1.0 weights entirely on target. The
default of 0.5 balances the weighting equally between data and target.
Try to better preserve local densities in the data.
Specifies whether the density-augmented objective of densMAP should be used for optimization.
Turning on this option generates an embedding where the local densities are encouraged to be
correlated with those in the original space.
Strength of local density preservation.
Controls the regularization weight of the density correlation term in densMAP. Higher values
prioritize density preservation over the UMAP objective, and vice versa for values closer to zero.
Setting this parameter to zero is equivalent to running the original UMAP algorithm.
Drop duplicate rows before embedding.
If you have more duplicates than you have n_neighbors you can have the identical data points lying
in different regions of your space. It also violates the definition of a metric. This option will
remove duplicates before embedding, and then map the original data points back to the reduced space. Duplicate
data points will be placed in the exact same location as the original data points.
