Missing data is (nearly) everywhere!

Missing Data in Time Series

Missing data is quite common, especially in sensor data.

Example pictures of measuring stations that produce time series
Reasons are manifold:
Data Recording, Data Transmission, Data Processing

The imputeTS package tries to help you!

Features of the package

drawing

  • Imputation
  • Visualization
  • Stats and Datasets


New Visualizations in Version 3.1

Getting an overview over your data.

Visualization of NA distribution

ggplot_na_distribution(tsAirgap)

Figure of plot_na_distribution plot

Digging deeper into missing data patterns.

Occurrence of Gap Sizes

ggplot_na_gapsize(tsNH4)

Figure of plot_na_gapsize plot

Exploring Imputation results.

Visualization of Imputations

imp <- na_kalman(tsAirgap)
ggplot_na_imputations(tsAirgap, imp)

Figure of plot_na_imputations plot

Easy to use.

Basically a ggplot object

Can be customized in the ggplot2 way 

ggplot_na_gapsize(tsNH4) +
theme(legend.position = "none")


Additionally direct parameters for typical plot adjustments 

ggplot_na_gapsize(tsNH4, legend = F )

Get in Touch

Thanks.