Numeric Distances

Distance definition on numerical column(s), like for instance Euclidean or Manhattan distance. Parameters for missing value handling and normalization can be set depending on the selected distance function.

Options

Column selection

Choose the columns for which the numeric distance measure is defined.

Distance Selection

Select and configure the distance to be used for measurement. Standard Distance (Euclidean/ Manhattan) The Minkowski or Lp-Norm distance generalizes the Euclidean or Manhattan distance Cosine distance Cosine similarity is a measure of similarity between two vectors of an inner product space that measures the cosine of the angle between them. This distance function is 1 - cosine-similarity and can take range [0,2].

Distance configuration

Choose the distance measure to use.

• Euclidean: Also known as the L₂ distance. It is the distance between 2 points which one would measure with a ruler.
• Manhattan: Also known as the L₁ distance. It is the sum of the absolute differences of the variable pairs.
• Maximum: Also known as the Chebyshev distance. It is the maximum of the absolute differences of the variable pairs.
• Custom: Free customization of the p parameter of the p-norm. A value less than 1 results in a function that is not a metric because it violates the triangle inequality.

P value

The power parameter for the L-norm distance. A value of 1 results in Manhattan distance, while a value of 2 equals Euclidean distance.

Normalize distance

If enabled, the distance will be normalized to the range [0, 1], where 1 refers to the greatest possible difference between the given input vectors. Note that this assumes that the input vectors are already normalized.

Missing value handling

Determines how missing values are handled during distance computation.

• Fail: Fails if a missing value cell occurs during computation
• Assume equal: Assume a missing value has the value of the respective counterpart - this will add 0 to the sum of pairwise absolute differences
• Average distance: Ignores the missing value and the corresponding value

Views

This node has no views

Workflows

• [03] F1 - Circuit GroupingsKNIME Hub
• 01_Anomaly_Detection_DemoKNIME Hub
• 01_Anomaly_Detection_DemoKNIME Hub
• 01_Anomaly_Detection_DemoKNIME Hub
• 01_Anomaly_Detection_DemoKNIME Hub
• Show all 346 workflows

Links

• No links available

Developers