Flat, Rich

REST API

{
  quote: {
    system: {
      site: {
        address_1: '564 Market St'
      }
    }
  }
}

Glossary

{
  solar_pmts_net_sum_yrs: 9155.44,
  solar_pmts_net_yr0: 0.0,
  solar_pmts_net_yr1: 610.36
  ...
}

django-vector

Self-documenting

How does it work?

apps/solar/product/metrics.py


class ProductNameMetric(FieldMetric):
    model = SolarProduct
    identifier = 'product_name'
    field = 'name'


class ProductNamesMetric(ListAggregationMetric):
    identifier = 'product_names'
    dependencies = [ProductNameMetric]

def _build_solarsite_dataframe(solarsites):
    metrics = [
        AddressLine1Metric,
        CityNameMetric,
        ContactNameMetric,
        SolarSiteLastUpdatedMetric,
        SolarSiteIDMetric,
        StateAbbreviationMetric,
        ZipcodeMetric
    ]

    return django_vector.build_dataframe(
        solarsites,
        metrics
    )

build_dataframe

+ build_single

Metric Types

  # django_vector/metrics.py


  class FieldMetric

  class ConcatenationMetric

  class ListAggregationMetric

More Metric Types

# coming soon: more aggregation

class AverageAggregationMetric
class CountMetric

# future work

class DateFieldChangedMetric

class DateQuoteSignedMetric(DateFieldChangedMetric):
   model = Quote
   identifier = 'date_quote_signed'
   field = 'status'
   value = 'SIGN'

Metric interface

class Metric():
    """Base class which defines the Metric interface."""

    dependencies = []

    def calculate(self, df):
        raise NotImplementedError

    def get_dependency_fields(self):
        raise NotImplementedError

Use cases

Sales & Ops Pipelines

Custom API endpoints

• When we calculate things like savings by default on the public api, things can get slow

"Hey customer can we turn off this field"