Numerical Timeseries Data
Timeseries data consists of measurements arranged in sequence at regular time intervals.
From Readings to a Sequence
Imagine a weather station recording conditions, waiting for the same amount of time, and recording them again. A single reading tells you about one moment. The complete ordered collection tells you how the measurements develop over time. That ordered collection is numerical timeseries data.
Timeseries data consists of measurements arranged in sequence at regular time intervals.
The word numerical emphasizes that the observations are measurements, such as temperature, pressure, or humidity. The word sequence emphasizes that their order matters. The word regular means that observations are collected at fixed intervals rather than at unrelated or unspecified times.
The Weather Dataset
The dataset comes from a weather station at the Max Planck Institute for Biogeochemistry in Jena, Germany. It contains 14 weather-related quantities recorded every 10 minutes. The original data reaches back to 2003, while this forecasting example uses data from 2009 through 2016.
Each time point can be viewed as a group of aligned measurements: the weather quantities recorded at that moment. Air temperature is therefore not detached from the rest of the record. It is one quantity associated with the same time as atmospheric pressure, humidity, wind direction, and the other weather measurements.
| Dataset feature | Description |
|---|---|
| Source | A weather station at the Max Planck Institute for Biogeochemistry in Jena, Germany |
| Quantities | 14 weather-related quantities |
| Recording interval | Every 10 minutes |
| Data used in the example | 2009 through 2016 |
| Examples of quantities | Air temperature, atmospheric pressure, humidity, and wind direction |
Important structural features of the weather dataset
History and Forecast Target
The forecasting task separates the data into two roles. A few days of recent weather data form the historical input supplied to the model. The air temperature 24 hours in the future is the forecast target that the model is asked to estimate.
Reading the Forecasting Setup
A forecasting task uses a few days of recent weather data to estimate air temperature 24 hours in the future. Identify the input and the target.
- 1Locate the historical window
The recent weather measurements from the few days before the forecast are the evidence supplied to the model.
- 2Locate the future point
The air temperature at the time 24 hours after the historical input is the value the model is asked to estimate.
- 3Keep the roles separate
The historical measurements are inputs, while the future air temperature is the target. They are not interchangeable parts of the task.
Input: a few days of recent weather data. Target: air temperature 24 hours in the future.
Why Order Matters
A weather record can be considered in relation to the records before and after it because the station produces observations in an ordered history. The measurements are not just isolated records with no connection. Each new observation follows an earlier observation after a fixed interval, and the accumulated history is used to frame a future prediction.
This is why sequence data is not limited to words in a sentence. Words have an order in a sentence, and weather measurements have an order in time. In both cases, the position of an item within the sequence helps describe what it means in relation to the other items.
What do you think happens?
In the temperature-forecasting task, which value is the forecast target?
Reveal answer
The air temperature 24 hours in the future.
The recent measurements are the input evidence. The future air temperature is the specific value the model is asked to estimate.
Mistakes in Reading the Task
Common mistake Treating the dataset as an unordered collection of weather records
What it looks likeDescribing a measurement without considering where it appears in the timeline
Why it failsTimeseries data is defined by measurements arranged in sequence at regular time intervals. The ordering is part of the information.
Do this insteadRead each measurement as belonging to a particular position in the ordered history.
Common mistake Calling every weather measurement the forecast target
What it looks likeSaying that pressure, humidity, temperature, and wind direction are all automatically the future target
Why it failsThe forecasting setup specifically asks for air temperature 24 hours in the future. The recent measurements serve as input evidence.
Do this insteadName the role of each part: recent weather data is the historical input, and future air temperature is the target.
Common mistake Confusing one time point with one isolated number
What it looks likeTreating a timestamp as if it contained only air temperature
Why it failsEach time point can be viewed as a group of aligned weather quantities recorded at that moment.
Do this insteadAssociate the timestamp with air temperature, atmospheric pressure, humidity, wind direction, and the other recorded quantities.
Common mistake Assuming sequence data means only text
What it looks likeRejecting weather measurements as sequence data because they are not words
Why it failsA weather station also produces a sequence by recording measurements at fixed intervals over time.
Do this insteadLook for ordered observations at regular intervals, whether the observations are words or numerical weather measurements.
Applying the Definition
Explain the weather forecasting problem in two parts. First, state what information is supplied as historical input. Then, state what future value is the target.
Hints
- The input covers a few days immediately before the forecast.
- The target is a specific measurement 24 hours later.
- Use the terms historical input and forecast target in your explanation.
A learner says, "The weather dataset is just a table of unrelated measurements." Correct the statement using the ideas of regular intervals, aligned quantities, and ordered history.
Hints
- Mention that observations are recorded every 10 minutes.
- Explain that several weather quantities belong to the same time point.
- Explain why earlier and later records form a sequence.
Key Takeaways
- 1Numerical timeseries data consists of measurements arranged in sequence at regular time intervals.
- 2The weather dataset contains 14 weather-related quantities recorded every 10 minutes at a station in Jena, Germany.
- 3The forecasting example uses weather data from 2009 through 2016, although the original data reaches back to 2003.
- 4A few days of recent weather measurements are the historical input, while air temperature 24 hours in the future is the forecast target.
- 5Weather measurements are sequence data because the station produces an ordered history in which each observation follows another after a fixed interval.
At a glance
- Numerical timeseries data is an ordered set of measurements collected at regular time intervals.
- Each weather time point groups several aligned quantities, including air temperature, atmospheric pressure, humidity, and wind direction.
- The forecasting task uses a few days of recent weather data as input and predicts air temperature 24 hours in the future.
- The order of weather observations matters because the station builds a history over time rather than producing unrelated isolated records.