🚩 Create a free WhyLabs account to get more value out of whylogs!

Did you know you can store, visualize, and monitor whylogs profiles with the WhyLabs Observability Platform? Sign up for a free WhyLabs account to leverage the power of whylogs and WhyLabs together!

Monitoring Regression Model Performance Metrics

Open in Colab

In this tutorial, we'll show how you can log performance metrics of your ML Model with whylogs, and how to send it to your dashboard at Whylabs Platform. We'll follow a regression use case, where we're predicting the temperature of a given location based on metereological features.

We will:

  • Download Weather Data for 7 days
  • Log daily input features with whylogs
  • Log daily regression performance metrics with whylogs
  • Write logged profiles to WhyLabs' dashboard
  • Show performance summary at WhyLabs

Installing whylogs

First, let's install whylogs. Since we want to write to WhyLabs, we'll install the whylabs extra. Additionally, we'll use the datasets module, so let's install it as well:

In [1]:
%pip install 'whylogs[whylabs, datasets]'

🌤️ The Data - Weather Forecast Dataset

In this example, we will use the Weather Forecast Dataset, using whylogs' Datasets module.

This dataset contains several meteorological features at a particular place (defined by latitude and longitude features) and time. The task is to predict the temperature based on the input features.

The original data is described in Shifts: A Dataset of Real Distributional Shift Across Multiple Large-Scale Tasks, by Malinin, Andrey, et al., and was further transformed to compose the current dataset. You can have more information about the resulting dataset and how to use it at https://whylogs.readthedocs.io/en/latest/datasets/weather.html.

Downloading the data into daily batches

Let's download 7 batches with 7 days worth of data, corresponding to the last 7 days. We can use directly the datasets module for that.

In [1]:
from whylogs.datasets import Weather
from datetime import datetime, timezone, timedelta
dataset = Weather()

start_timestamp = datetime.now(timezone.utc) - timedelta(days=7)
dataset.set_parameters(inference_start_timestamp=start_timestamp)

daily_batches = dataset.get_inference_data(number_batches=7)

#batches is an iterator, so let's get the list for this
daily_batches = list(daily_batches)

Since in this example we're mainly concerned with regression metrics, let's select a subset of the available features, for simplicity.

meta_climate, meta_latitude, meta_longitude will be our input features, while prediction_temperature is the predicted feature given by a trained ML model and the temperature feature is our target.

Let's take a look at the data for the first day:

In [2]:
columns = ["meta_climate","meta_latitude","meta_longitude","prediction_temperature","temperature"]
df = daily_batches[0].data[columns]

df.head()
Out[2]:
meta_climate meta_latitude meta_longitude prediction_temperature temperature
date
2022-09-07 00:00:00+00:00 mild temperate 38.891300 -6.821330 9.163181 10.0
2022-09-07 00:00:00+00:00 tropical 12.216667 109.216667 26.220221 27.0
2022-09-07 00:00:00+00:00 dry 37.991699 -101.746002 13.178478 15.0
2022-09-07 00:00:00+00:00 mild temperate -23.333599 -51.130100 23.255124 25.0
2022-09-07 00:00:00+00:00 mild temperate -23.479445 -52.012222 27.851674 32.0

✔️ Setting the Environment Variables

In order to send our profile to WhyLabs, let's first set up an account. You can skip this if you already have an account and a model set up.

We will need three pieces of information:

  • API token
  • Organization ID
  • Dataset ID (or model-id)

Go to https://whylabs.ai/free and grab a free account. You can follow along with the examples if you wish, but if you’re interested in only following this demonstration, you can go ahead and skip the quick start instructions.

After that, you’ll be prompted to create an API token. Once you create it, copy and store it locally. The second important information here is your org ID. Take note of it as well. After you get your API Token and Org ID, you can go to https://hub.whylabsapp.com/models to see your projects dashboard. You can create a new project and take note of it's ID (if it's a model project it will look like model-xxxx).

In [3]:
import getpass
import os

# set your org-id here - should be something like "org-xxxx"
print("Enter your WhyLabs Org ID") 
os.environ["WHYLABS_DEFAULT_ORG_ID"] = input()

# set your datased_id (or model_id) here - should be something like "model-xxxx"
print("Enter your WhyLabs Dataset ID")
os.environ["WHYLABS_DEFAULT_DATASET_ID"] = input()


# set your API key here
print("Enter your WhyLabs API key")
os.environ["WHYLABS_API_KEY"] = getpass.getpass()
print("Using API Key ID: ", os.environ["WHYLABS_API_KEY"][0:10])
Enter your WhyLabs Org ID
Enter your WhyLabs Dataset ID
Enter your WhyLabs API key
Using API Key ID:  z8fYdnQwHr

📊 Profiling the Data + Sending to WhyLabs

Traditionally, data is logged by calling why.log(). In this case, we'll use why.log_regression_metrics(), which will log data all the same, but additionally it will compute regression metrics and add them to your results.

log_regression_metrics takes the complete dataframe as input (with input/output features, as well as your prediction column and target column). We also have to define which column is our target (in this case, temperature) and which is our prediction column (prediction_temperature in this case).

Once the profile is logged, we can set it's timestamp for the proper day as given by our batch's timestamp.

Now that we have properly timestamped profiles with regression metrics, we can use the writer method to send it to WhyLabs:

In [4]:
from whylogs.api.writer.whylabs import WhyLabsWriter
import whylogs as why

columns = ["meta_climate","meta_latitude","meta_longitude","prediction_temperature","temperature"]
for batch in daily_batches:
    dataset_timestamp = batch.timestamp

    df = batch.data[columns]
    print("logging data for date {}".format(dataset_timestamp))
    results = why.log_regression_metrics(df, target_column = "temperature", prediction_column = "prediction_temperature")

    profile = results.profile()
    profile.set_dataset_timestamp(dataset_timestamp)
    print("writing profiles to whylabs...")
    results.writer("whylabs").write()
logging data for date 2022-09-07 00:00:00+00:00
writing profiles to whylabs...
logging data for date 2022-09-08 00:00:00+00:00
writing profiles to whylabs...
logging data for date 2022-09-09 00:00:00+00:00
writing profiles to whylabs...
logging data for date 2022-09-10 00:00:00+00:00
writing profiles to whylabs...
logging data for date 2022-09-11 00:00:00+00:00
writing profiles to whylabs...
logging data for date 2022-09-12 00:00:00+00:00
writing profiles to whylabs...
logging data for date 2022-09-13 00:00:00+00:00
writing profiles to whylabs...

And that's it! You just sent your profiles to WhyLabs.

At your model's dashboard, you should see the model metrics for the last seven days. For regression, the displayed metrics are:

  • Total output and input count
  • Mean Squared Error
  • Mean Absolute Error
  • Root Mean Squared Error

alt text