BigML Bindings: 101 - Using a Time SeriesΒΆ

Following the schema described in the prediction workflow, document, this is the code snippet that shows the minimal workflow to create a time series and produce a forecast.

from bigml.api import BigML
# step 0: creating a connection to the service (default credentials)
api = BigML()
# step 1: creating a source from the data in your local "data/grades.csv" file
source = api.create_source("data/grades.csv")
# waiting for the source to be finished. Results will be stored in `source`
# step 3: creating a dataset from the previously created `source`
dataset = api.create_dataset(source)
# waiting for the dataset to be finished
# step 5: creating a time series to forecast any numeric field in the dataset
time_series = api.create_time_series(dataset)
# waiting for the time series to be finished
api.ok(time_series) # finished time series
input_data = {"000005": {"horizon": 10}} # 10 points forecast for field ID 000005
forecast = api.create_forecast(time_series, {"000005": {"horizon": 10}})

In the previous code, the api.ok method is used to wait for the resource to be finished before calling the next create method or accessing the resource properties. In the first case, we could skip that api.ok`call because the next `create method would internally do the waiting when needed.

To learn more about the arguments that can be set in the forecast and timeseries creation calls and the response properties, please have a look at the API documentation

If you want to produce your forecasts locally, you can use the TimeSeries class in the timeseries module. A simple example of that is:

from bigml.timeseries import TimeSeries
local_time_series = TimeSeries("timeseries/5968ec46983efc21b000001b")
# 10 points forecast for field "Final" and 5 points for field "Assignment"
local_time_series.forecast({"Final": {"horizon": 10},
                            "Assignment": {"horizon": 5}})

Or you could store first your time series information in a file and use that file to create the local TimeSeries object:

# downloading the time series JSON to a local file
from bigml.api import BigML
api = BigML()
# creating the time series from the file
from bigml.timeseries import TimeSeries
local_time_series = TimeSeries("my_timeseries.json")
# 10 points forecast for field "Final" and 5 points for field "Assignment"
local_time_series.forecast({"Final": {"horizon": 10},
                            "Assignment": {"horizon": 5}})

Every modeling resource in BigML has its corresponding local class. Check the Local resources section of the documentation to learn more about them.