Forecasting Quick Start


This is a quick walkthrough of the basics of the using the Nexosis API. By following this walkthrough which uses a sample dataset, you will learn all of the steps needed to make forecasts using the Nexosis API.


Step 1: Prepare data for upload

In order to start using the Nexosis API, you’ll need to upload some data for the API to process. Data can be uploaded by posting the rows and columns as JSON, or, as a CSV file. We have a few datasets available which have examples of sales per day at a store, and sales of a single product per day. One of the sample datasets look like this:

timeStamp,sales,transactions
2012-12-31 00:00:00,2922.13,459
2013-01-01 00:00:00,1500.56,195
2013-01-02 00:00:00,4078.52,696
...

Since we are working with time series data, each row must have a timestamp of when the measurement occurred. The other columns are values which the Nexosis API can run algorithms against. These are the values which we are interested in forecasting.


Step 2: Start a Session

Now that we have some data, let’s upload it and get a forecast of how these values will change over time. To do this, we need to first send our data to a Dataset. Then we start a Session, referencing the Dataset we just created and containing parameters needed to determine how the Nexosis machine learning algorithms should work. Once the Session is started, our algorithms will start crunching the numbers to produce a set of forecast results.

For this dataset, we want to forecast the sales for the first quarter of 2017. All we need to do is specify the StartDate and EndDate as 2017-01-01 and 2017-04-01. The TargetColumn parameter also needs to be specified, which is the value which will be forecasted for this date range. We will set this value to sales.

Putting this all together, we will have a two requests that look like the ones below. Make sure to replace the {subscription key} section with your actual subscription key, and replace the file path with the path to one of the sample files that was downloaded earlier.

Upload a file

curl -v -X PUT "https://ml.nexosis.com/v1/data/location-a" \
             -H "Content-Type: text/csv" \
             -H "api-key: {subscription key}" \
             --data-binary "@/path/to/file/Location A.csv"

Start a Session

curl -v -X POST "https://ml.nexosis.com/v1/sessions/forecast?dataSetName=location-a&targetColumn=sales&startDate=2017-01-01&endDate=2017-03-31" \
             -H "Content-Type: application/json" \
             -H "api-key: {subscription key}" \

Once the session has been started, you should see a response similar to this:

{
  "sessionId": "{sessionId}",
  "type": "forecast",
  "status": "requested",
  "extraParameters": {},
  "dataSetName": "location-a",
  "targetColumn": "sales",
  "startDate": "2017-01-01T00:00:00+00:00",
  "endDate": "2017-04-01T00:00:00+00:00",
  "columns": 
  {
    "timeStamp": { dataType: "date", role: "timestamp" },
    "sales": { dataType: "numeric", role: "target" },
    "transactions": { dataType: "numeric" }
  }
}

Here we can see that we have a sessionId, which we will need later on. Also, the status of the session is now requested. The parameters that we sent up before are also echoed back to us. Now that we have requested a session, we can check the status to see when it completes by sending a GET with the sessionId we just got.

Check status of session

curl -v -X GET "https://ml.nexosis.com/v1/sessions/{sessionId}" \
            -H "api-key: {subscription key}"

Once this request comes back with a status of completed, the forecast will be available for download.


Step 3: Download results

Results can be downloaded by issuing a GET to the results endpoint.

Download session results

curl -v -X GET "https://ml.nexosis.com/v1/sessions/{sessionId}/results" \
            -H "api-key: {subscription key}"

The body of this response is the forecasted values over the requested date range. The results will be formatted like this.

{
  "data": [
    {
      "timestamp": "2017-01-01T00:00:00+00:00",
      "sales": 1911.46871429984
    },
    {
      "timestamp": "2017-01-02T00:00:00+00:00",
      "sales": 4330.20981465731
    },
    {
      "timestamp": "2017-01-03T00:00:00+00:00",
      "sales": 4573.98547777211
    },
...

The values object contains a key value pair of the column name, and the prediction value. You can see what these results look like by plotting both datasets in your favorite charting library.


Next steps

The Nexosis API can also do impact analysis of events. This can be used to gauge, for example, how impactful promotions or special events are on sales numbers. These types of sessions can be posted in the same way as forecasts, but to the sessions/impact endpoint.

Now that you are familiar with the basics, try running forecasts against new datasets, or, take a look at the code samples and client libraries, and write an application which integrates with the API. Show us what you were able to build!