imputeTS specializes in missing value imputation for univariate, equally-spaced numeric time series, with a focus on sensor and IoT data. It offers multiple imputation algorithms, ggplot2-based diagnostic plots for understanding and evaluating missing data patterns, and works naturally in tidy pipe workflows.
What’s covered#
- Mission, features, and scope – univariate numeric equally-spaced time series
- Imputation algorithms –
na_interpolation(),na_kalman(),na_locf(),na_ma(),na_mean(),na_random(),na_seadec(),na_seasplit(),na_replace(),na_remove() - Missing data overview plots –
ggplot_na_distribution(),ggplot_na_intervals(),ggplot_na_gapsize() - Imputation analysis plots –
ggplot_na_imputations()for evaluating imputation quality - Workflows – pipe-based usage with forecast packages
- Datasets –
tsAirgap,tsNH4,tsHeating