sample-catalog-item.yml

# Sample catalog item entry

# note: the filename is catalog item's  unique identifier (which should have a .yml file extension)
# only lowercase characters, no special characters, underscore or spaces.
# only the dash '-' is allowed in the filename

# short name of catalog item
name: 'Sample Catalog Item Entry (required)'

# description can be as long as you like
description: 'Description of catalog item (required)'

# card-type (required) can be 'model' or 'dataset'
card-type: model

# organization providing the catalog item (required)
organization:
  name: 'Name of your organization'
  url: 'https://www.example.com'

# date catalog item was added/created (required)
date-added: '2022-03-02'

# date catalog item was modified after added/created (optional)
date-modified: '2022-03-20'

# tags which can be associated with your dataset (optional)
# lowercase characters, no special characters, underscore or spaces, use dash (-) allowed
tags:
  - relative-wealth
  - timor-leste
  - sklearn-model
  - machine-learning

# country or region to which your catalog item applies (optional, can have multiple values)
# if supplied, each entry must be in the list in folder validation/regions.txt
country-region: timor-leste
# if multiple countries/regions are supplied
# alternate format
# country-region:
#   - timor-leste
#   - malaysia
#   - thailand
#   - myanmar

# year or time period to which your catalog applies (optional)
year-period: 2016

# evaluation metrics (optional) - can either be value and metric, or a link
# for further discussion, or both.
evaluation-metrics:
  - metric:
      metric-type: rsquared
      value: 0.59
  - link:
      description: 'Additional evaluation metrics discussion'
      url: https://thinkingmachines.github.io/unicef-ai4d-poverty-mapping/notebooks/2023-01-17-initial-model-ph-mm-tl-kh/model_tl.html#evaluate-model-training-using-cross-validation

# links are the URLs to the datasets, model weights, documentation, or github repos
# the valid types are 'model-weights','dataset-geojson', 'dataset-csv',
# 'code-notebook','repository', 'documentation', 'research-paper', 'limitations',
# 'training-dataset-csv','training-dataset-geojson'
# at least one link must be provided
links:
  - url: https://github.com/butchtm/unicef-ai4d-poverty-mapping/blob/feat/convert-geojsons-to-csvs/notebooks/2023-02-21-single-country-rollouts/model_tl_rollout.pkl
    description: 'Timor Leste Relative Wealth Model Weights (.pkl). See related notebooks for usage'
    type: model-weights

    # dataset links are those that have a type that start with keyword "dataset"
    # and must have a name - but depending on the url, the name can be derived from the last part of the url

    # if the dataset is a csv type (has a "csv" in the type), we strongly recommend adding HXL Tags.
    # Please visit https://hxlstandard.org/ to learn how to add HXL tags to your datasets.
  - url: https://drive.google.com/file/d/15xrSAyWkFZRwVuGtECTRwNz06amRsjew/view?usp=share_link
    description: 'Training dataset for Relative Wealth Model on Timor Leste as CSV'
    type: 'training-dataset-csv'
    name: 2023-02-21-tl-dhs-cluster-features-all.csv
    # skip-hxl-tag-validation (optional) is a boolean flag that will skip hxl tag validation. the default value is false and is false if not present
    # the hxl-tag-validation is only applied to datasets with a `csv` type
    # skip-hxl-tag-validation: true

  - url: https://github.com/butchtm/unicef-ai4d-poverty-mapping/blob/feat/convert-geojsons-to-csvs/notebooks/2023-02-21-single-country-rollouts/rollout_output_tl.geojson
    description: 'Country wide rollout (inference) of the Relative Wealth Model on Timor Leste using 2.4km grids (bingtile quadkey level 14) as geojson'
    type: 'dataset-geojson'
    # name is optional for dataset links if it can be derived from the url
    # name: rollout_output_tl.geojson

    # alt-format (optional): if the dataset provides multiple formats (csv, geojson, shapefile) - they can be specified as alternate formats
    # multiple times, each alt-format providing a separate url and type
    # if the dataset is a csv type, hxl-tag validation is applied unless the skip-hxl-tag-validation flag is set to true
    alt-format:
      - url: https://drive.google.com/file/d/138zqndK07bqtUXAelzpiFqacohVxWetx/view?usp=share_link
        type: 'dataset-csv'
        # skip-hxl-tag-validation (optional) is a boolean flag that will skip hxl tag validation. the default value is false and is false if not present
        # the hxl-tag-validation is only applied to datasets with a `csv` type