Custom Metrics


Custom metrics offer the flexibility to define metrics that align precisely with your machine learning requirements. Whether it's tracking business KPIs, crafting specialized performance assessments, or computing weighted averages, custom metrics empower you to tailor measurements to your specific needs. Seamlessly integrate these custom metrics throughout the Fiddler platform, leveraging them in dashboards, alerting, and performance tracking.

Craft novel metrics by harnessing a blend of your model's existing dimensions and functions, employing a familiar query language called Fiddler Query Language (FQL) . This capability enables you to amalgamate features, metadata, predictions, and actual outcomes using a rich array of aggregations, operators, and metric functions, thereby expanding the depth of your analytical insights.

How to Define a Custom Metric

Craft bespoke metrics effortlessly with Fiddler's intuitive Excel-formula-like syntax. Once a custom metrics is defined, Fiddler distinguishes itself by seamlessly managing time granularity and ranges within the platform's charting, dashboarding and analytics experience. This empowers you to effortlessly adjust time ranges and granularities without the need to modify your query, ensuring a smooth and efficient analytical experience.

Fiddler Custom Metrics are constructed using the Fiddler Query Language (FQL).

πŸ“˜ Custom metrics must return either:

  1. an aggregate (produced by aggregate functions or built-in metric functions), or

  2. a combination of aggregates

Let us illustrate further by providing a few examples.



Let’s say you wanted to create a custom metric for the following:

If an event is a false negative, assign a value of -40. If the event is a false positive, assign a value of -400. If the event is a true positive or true negative, then assign a value of 250.

We can formulate this metric using FQL with the following code:

average(if(fn(), -40, if(fp(), -400, 250)))

Fiddler offers convenience functions, like fp() and fn(). Alternatively, we could also identify false positives and false negatives the old fashioned way.

average(if(Prediction < 0.5 and Target == 1, -40, if(Prediction >= 0.5 and Target == 0, -400, 250)))

Here, we assume Prediction is the name of the output column for a binary classifier and Target is the name of our label column

Tweedie Loss

In our next example, we provide an example implementation of the Tweedie Loss Function. Here, Target is the name of the target column and Prediction is the name of the prediction/output column.

average((Target \* Prediction ^ (1 - 0.5)) / (1 - 0.5) + Prediction ^ (2 - 0.5) / (2 - 0.5))

Adding a Custom Metric

To learn more about adding a Custom Metric using the Python client, see fdl.CustomMetric


PROJECT = fdl.Project.from_name(name=PROJECT_NAME)
MODEL = fdl.Model.from_name(name=MODEL_NAME,

METRIC = fdl.CustomMetric(
        definition="average(if(\"spend_amount\">1000, \"spend_amount\", 0))", #Use Fiddler Query Languge (FQL) to define your custom metrics
        description='Get average spend for users spending over $1000',

Modifying Custom Metrics

Since alerts can be set on Custom Metrics, making modifications to a metric may introduce inconsistencies in alerts.

🚧 Therefore, custom metrics cannot be modified once they are created.

If you'd like to try out a new metric, you can create a new one with a different name and definition.

Last updated

Β© 2024 Fiddler AI