Skip to content

Latest commit

 

History

History
154 lines (112 loc) · 5.51 KB

dark_matters_rnd.md

File metadata and controls

154 lines (112 loc) · 5.51 KB
Raw

Dark Matters R&D Info

Only supports once jobs right now

This datagen processor will generate information about this fictional company's R&D projects. Specifically created for the CIS 405 team project, it will generate BI records with 2 key facts and 3 key dimensions (technically 4 if you include the project_uuid). Each record represents a specific R&D project (assuming a unique uuid in the dataset), and will contain the following information:

  • project_uuid: Project ID. Will be determined randomly to differentiate project records and is limited to alphanumericals for compatibility with Excel.
  • project_cost (fact 1): Overall cost (in $1k's) of the R&D project including labor, materials, production, etc.
  • savings (fact 2): Overall reduction of manufacturing, maintenance, and operational expenses for our next 100 launches in $1k's
  • location (dimension 1): Facility location for the R&D project by US state (This generator assumes you provided a valid array of two-digit state names and will break if not...)
  • funding_agency (dimension 2): Agency/organization funding the R&D project
  • subsystem (dimension 3): List of subsystems the R&D projects will span

Options

size

Valid Options Default Required
Any number your computer can handle 1000 N

Overall size of the dataset since this only supports once jobs. Minimum record count for the class is 50, but 500-1k should likely be enough to get meaningful trends in the data.

cost_max

Valid Options Default Required
Any number equal to or greater than 1 1 N

Maximum R&D project cost in $1k.

savings_factor_max

Valid Options Default Required
Any number equal to or greater than 1 1 N

Cost savings factor (including manufacturing, maintenance, and operational expenses) for the next 100 launches as a direct result of the particular R&D project. Will be multiplied the cost_max parameter, so this should probably be something less than 10 or so (10 would be a 10x saving for a given project cost)

uuid_length

Valid Options Default Required
Any positive number 7 N

This will be passed through Math.floor() to determine the project uuid length

locations

Valid Options Default Required
Array of state abbreviations ['AZ', 'CA', 'WA', 'FL'] N

Array of facility locations by state in two-letter format. There's no filtering right now, so invalid state abbreviations WILL RESULT IN A BROKEN DATASET.

funding_agencies

Valid Options Default Required
Any array ['NASA', 'DoD', 'DoE'] N

This is a free market, so any array of entities/organizations will be allowed. Must include at least one funding source since money does not randomly appear for projects.

subsystems

Valid Options Default Required
Any array ['propulsion', 'avionics', 'comms'] N

Pretty much just a list of rocket subsystems or subcomponents.

formatted_field_names

Valid Options Default Required
Boolean false N

Determines whether or not to lowercase the field names and replace spaces with _

biasing

Valid Options Default Required
Object describing biasing relations '' Y

This isn't filtered right now, so invalid configs will result in broken jobs. With the following example, a funding agency will be randomly selected and used to alter the probability of selecting different locations. (biasing_probability influences the likeliness of dimension biasing, so a 0 means there will be no dimension biasing, and a 1 means that there will always be dimension biasing.) After randomly generating a project cost and project savings, the location will be used to either increase or decrease the cost, and the subsystem will be used to increase or decrease the savings.

{
    "biasing_dimension": "funding_agencies",
    "biased_dimension": "locations",
    "biasing_probability": .5,
    "fact_biasing": {
        "cost": "locations",
        "savings": "subsystems"
    }
}

Sample Job

{
  "name": "Dark Matters RnD Generator",
  "lifecycle": "once",
  "workers": 1,
  "operations": [
    {
      "_op": "dark_matters_rnd",
      "size": 10000,
      "cost_max": 500,
      "savings_factor_max": 6,
      "locations": ["AZ", "CA", "WA", "FL", "MI", "TX", "NV", "NM"]
      "funding_agencies": ["ASU", "UofA", "DoD", "Internal", "DoA", "UAE", "NASA", "JAXA"]
      "subsystems": ["propulsion", "avionics", "comms", "mission ops", "stage separation", "payload", "fuel tanks", "recovery"],
      "biasing": {
          "biasing_dimension": "funding_agencies",
          "biased_dimension": "locations",
          "biasing_probability": .5,
          "fact_biasing": {
              "cost": "locations",
              "savings": "subsystems"
          }
      }
    },
    {
      "_op": "stdout"
    }
  ],
  "assets": [
      "classets"
  ]
}

Sample Record

Paired with the file-assets, this processor could be used to generate a CSV dataset.

{
    "Project ID": "5a414a2"
    "Project Cost": 385,
    "Project Savings": 1925,
    "Location": "AZ",
    "Funding Agency": "ASU",
    "Subsystem": "propulsion"
}