Table of contents

Installation, requirements and versions

Lists system requirements, installation instructions, Python and Modeling Assistant versions and Community Edition limitations.

System Requirements

IBM® Decision Optimization for Watson Studio was previously known as IBM Decision Optimization for Data Science Experience.

To find details of the hardware and software configurations required for installation of IBM Watson Studio refer to the system requirements page.

Installation

Note: By default, the Watson Studio installer also installs Decision Optimization for Watson Studio Community Edition. The engines included in this default installation are limited to solving problems of a certain size. To use the full capacity Decision Optimization engines and solve larger problems, you need to install an add-on. See Decision Optimization for Watson Studio for pricing.

To install Watson Studio follow the detailed step-by-step installation instructions provided for Watson Studio.

Select the appropriate version of the Decision Optimization add-on, according to your installed version of Watson Studio, as shown in the following table.

Table 1. Compatible versions
Watson Studio (DSX Local) Version Decision Optimization for Watson Studio (Data Science Experience) Installer
Watson Studio V1.2.2 and later Decision Optimization for Watson Studio V1.2.2 and later
DSX Local V1.2.1 Decision Optimization for Data Science Experience V1.1
DSX Local V1.2 Fix Pack 3 Decision Optimization for Data Science Experience V1.0.0.1 from Fix Central
DSX Local V1.2 Decision Optimization for Data Science Experience V1.0
To install the Decision Optimization full capacity add-on on top of an existing installation of Watson Studio:
  • Login as root on the master node of Watson Studio installation
  • Download the installation package TAR file into any directory
  • Run:
    tar xzvf <add-on-installer>.tar.gz
     ./install.sh <namespace where Watson Studio is installed, the default is 'dsx'>
The script will install the add-on on top of the existing Watson Studio installation. The installation log file will be saved as dods_addon_install.log in the current directory.

To uninstall the add-on, run ./uninstall.sh from the same directory.

For Watson Studio on IBM Cloud Private you can either use the previous installation method or the following alternative installation method:

If you installed Watson Studio on IBM Cloud Private (as documented in Install Watson Studio on the ICP catalog), you can also add the Decision Optimization add-on to the IBM Cloud Private catalog and install it from the catalog.

To add the Decision Optimization to the IBM Cloud Private catalog, complete the following steps:

  • Enter the following command to ensure that bx pr and docker are authenticated:
    bx pr login -a https://<cluster_ip>:8443 --skip-ssl-validation 
    docker login <cluster_name>:8500
  • In the IBM Cloud Private CLI tool, enter the following command:
    bx pr load-ppa-archive --archive <add-on-installer>.tar.gz
  • Set the Watson Studio images scope to global by entering the following command with kubectl authenticated (you must have administrator rights to do this):
    for image in dd-install dd-init dd-processor dd-processor-docplex-python-2; do kubectl get image $image -o yaml | sed 's/scope: namespace/scope: global/' | kubectl apply -f -; done

You can then install the Decision Optimization add-on from the catalog as follows:

In IBM Cloud Private administration UI:
  • Go to Catalog > Helm Charts and verify that the ibm-dods chart now displays.
  • Click Configure, enter a release name and the namespace where Watson Studio is installed.
  • Click Install to install the add-on

Upgrading from earlier versions

To upgrade from earlier versions of Watson Studio and Decision Optimization for Watson Studio:

  1. Upgrade Watson Studio
  2. Install Decision Optimization for Watson Studio on top of the migrated cluster.

    For example, to upgrade from DSX Local V1.2.0.3 with Decision Optimization for Data Science Local V1.2.1, to version V1.2.1, first upgrade DSX Local to V1.2.1 and then install Decision Optimization for Data Science V1.1.

Python version

The Decision Optimization for Watson Studio environment currently supports Python V2.7.

Problem size

As a Watson Studio user, you have free access to the Community Edition of the Decision Optimization engines. The Community Edition is limited to 1000 constraints and variables, and a search space of 1000 X 1000 for constraint programming problem. You can run optimization models that don't exceed this size without any additional configuration. To solve larger problems, install the Decision Optimization full capacity add-on as described above.

Modeling Assistant model types

The Modeling Assistant currently supports Resource Assignment, Scheduling, Supply & Demand Planning and Selection & Allocation models. See Selecting a Decision domain in the Modeling Assistant.