About OnePart Reduce

This introduces NETVIBES OnePart Reduce concepts and terminology

This page discusses:

What is OnePart Reduce?

In addition to exploring data in standard NETVIBES OnePart interfaces, OnePart Reduce allows you to:

  • Create Universes and define its clusters
  • Compare & qualify parts as STANDARDS
  • Publish clusters to the NETVIBES OnePart library
  • Classify parts
    Note: In addition to NETVIBES OnePart Standard license, you must have the OnePart licenses for the number of named users and documents you require.

How to access OnePart Reduce

This section details how to access OnePart Reduce. A dedicated OnePart Reduce home page is available from the home menu.



OnePart Reduce requires a special license. Users that have the proper rights and license can access Reduce from the OnePart UI as follows:

  1. Open OnePart in your web browser at http://<ONEPART_SERVERNAME>:<BASEPORT>/reduce
    1. For example, http://localhost:10000/reduce
  2. Log in to OnePart. You can either:
    1. Enter your local Windows user name and password. OR
    2. Enter the demo login User: user1 and Password: user1
      Note: The OnePart Reduce user must have access to all the files for clustering.
  3. This brings you to the OnePart Reduce home page.

Limited version

A limited version of OnePart Reduce includes a 2-user license and 15,000 imported items but does NOT include the publishing feature. Users can only view 1 out of 4 Clusters. Note that this evaluation mode is not available in the 3DDashboard. It is available with OnePart Standard at: http://<ONEPART_SERVERNAME>:<BASEPORT>/reduce

For more details, see Reduce evaluation mode

Document tokens

This section details how OnePart counts the document tokens.

The number of documents that can be processed in OnePart Reduce is restricted. To know how many documents you can use in Reduce, divide the "onepart.reduce.tokens" license feature value (default is 900) by the value of the "indexing.costs.business-item" license feature (default value is 0.06). So, by default, you can reduce 900/0.06 = 15000 documents.

Each document is counted only once (based on its URI) even if it is used in several universes. Even if you delete the universes containing a document, it will still be counted forever. When building a new universe, only new documents (never used in another universe) will be counted. It means that, even if there are no more reduce document tokens, you can still use OnePart Reduce.

When the limit is reached, some documents of the scope may be ignored. A warning ("i" icon turns red) on the universe will inform you of this.

Use case

Let's see how document tokens are counted. Consider that the licence we have authorized 15000 documents.

  1. We create a first universe A1 with 10000 documents, it will consume 10000 docs in Reduce (still 5000 available)
  2. We create a second universe A2 with the exact same 10000 documents (with different parameters), it will not consume anymore docs (still 5000 available).
  3. We create a third universe B with 3000 docs and within this 3000 docs there are 2000 docs that are in universe A1 (and A2); in this case it will consume 1000 docs (4000 available).
  4. We create a fourth universe C with 7000 docs ( 2000 of these docs are in A1, A2 or B); in this case it would consume 5000 new docs but you just have only 4000 available. There fore in the universe only 6000 docs will be processed and 1000 will be ignored and mentioned in the details of the universe.

Common terms used

Below are common terms used in OnePart Reduce.

  • Cluster is a group of similar parts.
  • Standard/alternate parts - in the standardization process, you define the Standard parts for each cluster. All other parts in the cluster are tagged as Alternates when published.
  • Universe is a group of clusters.
  • Facets are used to narrow search results. Use them to drill-down into an area, such as language, author, or file type.

Understanding the Reduce homepage

From the OnePart Reduce homepage, you can see the work done in a universe and the stats on all the data published in OnePart.

No. Section Detail
1 Universe statistics Displays statistics of potential duplicates on the entire dataset.
2 Run button Click this button to compute clustering on the entire dataset.
3

Parts status

Displays statistics by parts status including those parts flagged as masters, duplicates, unique or not yet published.

4

Create universe button

Click to create the following universe types:

1. Standardization - Eliminate duplicate parts by setting the standards in the universe; add classification - see Use case - Standardization

2. Analysis - You can either analyze:

  • an assembly to check if some parts could be replaced - see Use case - Analysis
  • 2 lists of parts to check if some parts could be replaced

See Clustering parameters for the universes

5 Cluster dashboard

See Cluster Dashboard

Cluster Dashboard

Cluster dashboard details for OnePart Reduce are below.

No.

Section

Detail

1

Universe link

Brings you to the Universe Clusters Overview.

See Understanding the Universe Clusters Overview

2 Information button Click this icon to see the detailed parameters of the universe including a link for the defined universe scope and related links.
3 Update button Updates the universe clusters with new parts within the scope now available in the OnePart index.

Note:

You cannot update a universe if there are published clusters with pending modifications

4

Delete button This deletes the universe and all the annotations done in this universe will be unpublished in OnePart.
5 Stats chart Related statistics display here.

For example, the potential alternates parts in this universe.

6 Cluster status

See Cluster Status.

Cluster Status

The cluster lifecycle in OnePart Reduce contains 4 stages detailed below. Status percentages display for the universe clusters.

  • Pending - these are the clusters pending modifications
  • Modified - these cluster has been modified
  • Validated - when the user clicks “Validated”
  • Published - After a cluster has been published, it can be updated by clicking Publish again or it can be unpublished

For example, only 2% of my part library is published and 96% is waiting for review. See also Editing Clusters.



Clustering parameters

When creating a universe in OnePart Reduce, you must set the clustering parameters. These determine how to split the clusters, set display attributes.

No. Section Detail
1 Similarity inside a cluster Determines the shape similarity strictness. By default, the slider is set to « medium ». This creates the clusters based on shape.
Note:

If you select a "Fuzzy" clustering you might get only partial similarity between the parts.

2 Taxonomy used for Categorization You can select a CSV file to apply a taxonomy used for the categorization universes. See Set up Taxonomy
3 Sub-clustering by attributes To split the shape clusters by either Material or Size:
  • It can be text (like Material) or it can be numerical values (like Size).
  • For numerical values, a threshold can be set (in absolute value or in %).
For example, to create clusters based only on shape, disable the Size attribute.

Note:

If certain parts in the cluster have no material value, Reduce will create an additional cluster.

4 Warning attributes in clusters

You can setup warnings on clusters by attributes

- see Set up Cluster warnings

Understanding the Universe Clusters Overview

For each universe, an overview page is available showing the clusters for the scope selection.

No. Section Detail
1 Cluster title The cluster name and number of clusters found.
2 Publish

this universe in OnePart

Publish all the clusters at once with the Validated status in OnePart.
3 Merge button Merge the selected clusters using the check boxes.
4 Switch view Switch to table view. For standardization universes, you can add classification to multiple parts directly in the table view.

5

Cluster maturity

Filter the universe clusters based on status. See also Cluster Status.

6

Cluster size

Filter the clusters based on the minimum cluster size.

7 Warning Filter the clusters by warnings, Classification, or by custom attributes.
8 Sub-clustering Filter by sub-clustering attributes such as Material.
9

Sort clusters by

For the Assembly analysis universe type, you can sort by the number of instances in the assembly.

For all clusters, you can sort by the number of parts in a cluster.

10

Cluster display

You can analyze the Universe clusters displayed here. By clicking on a cluster, you can explore it.

Editing Clusters

In an Standardization universe, you can choose which part should be kept as the standard, and all the others should be tagged as alternates or "not to be used".

You can also add classification in this page.



Use this area to:

No. Section Detail
1 Classification In a standardization universe, you can add classification names to the parts in the cluster. You can add as many sub-classification as you want. Enter Classification/SubClassification. For example, Wheels/Small-wheels. See Use case - Standardization
2 Cluster lifecycle +

Publish button

This is the cluster lifecycle and publish the clusters with the “Validated” status in OnePart.     

Note: If you want to publish all your validated clusters in the universe at once rather than cluster by cluster, use the "Publish this universe in OnePart" button  - Understanding the Universe Clusters Overview                  

3 Cluster toolbar Buttons include:
  • Show modifications -

    For published clusters that currently have pending modifications

  • Check all - to check all parts in the cluster to compare
  • Compare - to compare selected parts
  • Switch view - to switch to the table view Using the Table view
4 Cluster parts Displays all the parts in current cluster. You can:

  • Tag a part as the standard with the icon.

    The last part selected as a standard is used as the reference item and displays bold. The reference items is used for the 3D similarity comparison and displayed as the main cluster part.

  • Tag alternate parts with the icon.

  • Tag parts that should not be used with the icon.

  • Tag parts to be compared with the icon.

  • Bookmark parts for later with the icon.

    See also Merging clusters and Split cluster

Using the Table view

OnePart Reduce provides a table view of the parts in the cluster.



The table allows you to:

  • Set parts as standards, alternates or not for use
  • Pin columns to the left or right; Right-click on the column to pin it
  • Sort part by columns; Ctrl + Click to select multiple columns in the order you need
  • Select the columns to display

Comparing parts

You can compare the part’s attributes to help you make the right business decision. OnePart Reduce provides 2D comparison.

Note: As of OnePart R2019x FD03, the 3D Compare feature providing an exact 3D comparison for 2 selected parts is deprecated. You can still enable the feature from the Mashup Builder page with the proper license.

The standard OnePart 2D compare is available for 3 or more parts.



You can compare parts as follows. For example, let's take a look at the demo skateboard assembly.

  1. Click Check All or hover over each part and click to compare.

  2. Click Compare to view the part’s attributes.

  3. Examine the part’s attributes to help you make the business decision. Differences display in red.

Splitting parts from cluster

OnePart Reduce allows you to split parts from In this cluster.

For example, you might want only want the gray CAMs so you can split the two pink ones from the cluster. For assembly analysis, note that you cannot split parts from the cluster that are in the current assembly instance.

  • Click on the parts to split.
  • Click the blue Split parts button to split the parts from the cluster.


Merging clusters

OnePart Reduce allows you to merger clusters that display In other similar clusters with the parts In this cluster. The number in the blue square represents the number of parts in the cluster. When you merge, it will automatically add all the parts from the selected clusters to the current cluster. By clicking on the image, you can open the cluster to analyze it here. The similarity rules used are those defined in the user preferences. You can edit the 3D similarity preferences to directly view more or fewer clusters.

  1. Click on the cluster to merge.
  2. Click the blue Merge clusters button to merge the 2 clusters.
  3. Now you can tag the parts just added - see Editing Clusters