In andreas50/uts: Unevenly-spaced Time Series

Introduction

An Unevenly-spaced Time Series (uts) is a sequence of observation time and value pairs (tn, Xn) with strictly increasing observation times. As opposed to equally spaced time series, the spacing of observation times may not be constant.

Comparison with other packages for irregular time series

| Time series operation | In terms of number of observations | In terms of time window | |---------------------------------|------------------------------------|-------------------------| | Get initial subperiod | head | head_t | | Get terminal subperiod | tail | tail_t | | Delay observations | lag | lag_t | | Calculate differences over time | diff | diff_t | | Get the length | length | length_t |

Methods that operate in terms of the number of observations typically specify that number using an integer argument.

On the other hand, methods that operate in terms of a time window typically have an argument specifying the temporal width (also know as duration) of this time window[^1]. The number of observations falling into this window depends on the observation density of the time series.

[^1]: length and length_t are exceptions. Instead of requiring an integer or duration argument, respectively, they return such a value.

Indexing

Recycling rules

R uses a variety of rules for handling arithmetic between vectors and arrays of different lengths. These are commonly known as recycling rules, see Section 5.4.1 of "An introduction to R" for a complete list.

Arithmetic with uts objects only keeps a few rules that do not make strong implicit assumptions about what exactly the user wants to do. Specifically: