Multivariate Time Series Explorer
Dataset: Tongue Twister
Each graph below the topmost graph shows a single transformed variable graphed against time, with each time series corresponding to a single subject in the dataset. The transformation is Y'=weight*transformation(Y+shift) where weight, transformation and shift are specified for each graph in turn in the box to the right.
The topmost graph is a linear combination of the transformed underlying variables: a time series for subject X, say, in the topmost graph is the sum of all time series for subject X in the graphs below it.
Each time series must, for each variable, provide an observation at each time point over the range of time points that it covers. Missing observations have been added into the data set using straight line interpolation.
Hover over a point to obtain information about it.
Click on a point to highlight or unhighlight the points in the time series to which it belongs.
Sum of Y_{i} VS Time
Sum of Y_{i} | |
Time (Minutes) |
Y_{1}=weight*transformation(Witch+shift) VS Time
Y_{1} (Recitations) | |
Time (Minutes) |
Y_{2}=weight*transformation(Wristwatch+shift) VS Time
Y_{2} (Recitations) | |
Time (Minutes) |
Y_{3}=weight*transformation(Shipshape+shift) VS Time
Y_{3} (Recitations) | |
Time (Minutes) |