TSBenchmarks

Time-series benchmarking a service. TSBenchmarks is an SDK for benchmarking models on public datasets.

Since we take care of the whole infrastructure, benchmarking becomes as easy as running a line in your python nootebooks or calling an API.

Why?

We build TSBenchmarks because we wanted a standarized solution for benchmarking time-series forecasting models. We evaluate provided forecasts on well known public datasets against benchmark models.

Table of contents

• TSBenchmarks
  • Install
  • How to use
    • Data Format
    • 1. Request free trial
    • 2. Run tsbenchmarks on a private S3 Bucket
      • 2.1 Upload to S3 from python
      • 2.2 Run the evaluation process
      • 2.3 Download your results from s3

Available Benchmark Datasets

• M4-Daily

Available Benchmark models

• Naive1
• Naive2
• ETS
• Theta
• ARIMA
• MLP
• RNN
• ESRNN
• FFORMA
• NBEATS

Available Metrics and Plots

• MASE
• sMAPE
• Average loss by time-series
• Average loss by timestamp

Install

pip install tsbenchmarks

How to use

You can use TSBenchmarks by either using a completely public S3 bucket or by uploading a file to your own S3 bucket provided by us.

Data Format

Currently we only support .csv files. These files must include at least 3 columns, with a unique_id (identifier of each time-series) a date stamp and a value. The unique_id and ds must coincide with the test set of the selected benchmark dataset.

1. Request free trial

Request a free trial sending an email to: fede.garza.ramirez@gmail.com and get your BUCKET_NAME, API_ID and API_KEY, AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY.

2. Run tsbenchmarks on a private S3 Bucket

If you don´t want other people to potentially have access to your data you can run tsbenchmarks on a private S3 Bucket. For that you have to upload your data to a private S3 Bucket that we will provide for you; you can do this inside of python.

2.1 Upload to S3 from python

You will need the AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY that we provided.

• Import and Instantiate TSBenchmarks introducing your BUCKET_NAME, API_ID and API_KEY, AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY.

from tsbenchmarks.core import TSBenchmarks

tsbenchmarks = TSBenchmarks(bucket_name='<BUCKET_NAME>',
                            api_id='<API_ID>',
                            api_key='<API_KEY>',
                            aws_access_key_id='<AWS_ACCESS_KEY_ID>',
                            aws_secret_access_key='<AWS_SECRET_ACCESS_KEY>')

• Upload your local file introducing its name.

s3_uri = tsbenchmarks.upload_to_s3(file='<YOUR FILE NAME>')

• Run the evaluation process

To run the process specify:

• s3_uri: S3 uri provided after calling tsbenchmarks.upload_to_s3().
• dataset: Name of dataset
• ds_column: Name of the unique id column.
• y_column: Name of the target column.

#Run Evaluation
response_tmp_ft = tsbenchmarks.evaluate_my_model(
                    s3_uri="<PRIVATE S3 URI HERE>",
                    dataset='M4-Daily',
                    unique_id_column="<NAME OF ID COLUMN>",
                    ds_column= "<NAME OF DATESTAMP COLUMN>",
                    y_column="<NAME OF TARGET COLUMN>"
                  )

response_tmp_ft

|

• Monitor the process with the following code. Once it's done, access to your bucket to download the generated features.

job_id = response_tmp_ft['id_job'].item()

tsbenchmarks.get_status(job_id)

2.2 Download your results from s3

Once the process is done you can explore and download the results from s3.

ToDos

• Optimizing writing and reading speed with Parquet files
• Nan Handling
• Check data integrity before Upload
• Informative error messages
• Informative Status