Overview of time series (Series) objects
Time series objects are two- or higher-dimensional arrays whose rows are
referenced by dates.
The arrays can be one of the following types:
- numeric arrays
- string arrays
- cell arrays
Most of the time series methods are specifically designed for numeric time series,
and do not, naturally, work for string time series or cell time series.
Categorical list of functions
Constructing time series objects
Generating new time series from existing time series
Function |
Description |
Series.grow |
Cumulate level time series from differences or rates of growth |
Manipulating the time dimension
Function |
Description |
clip |
Clip time series to a shorter range |
rebase |
Rebase times series data to specified period |
Filtering, interpolating and aggregating time series
Function |
Description |
arf |
Create autoregressive time series from input data |
convert |
Convert time series to another frequency |
fillMissing |
Fill missing time series observations |
hpf |
Hodrick-Prescott filter with conditioning information |
moving |
Apply function to moving window of time series observations |
chainlink |
Calculate chain linked aggregate level series from level components and weights |
Statistics and regression
The following standard Matlab functions work on time series objects:
Function |
Default dimension |
Default output |
mean |
1 |
array |
median |
1 |
array |
mode |
1 |
array |
geomean |
1 |
array |
sum |
1 |
array |
prod |
1 |
array |
std |
1 |
array |
var |
1 |
array |
cov |
1 |
array |
prctile |
2 |
Series |
Function |
Description |
rmse |
Calculate RMSE for given observations and predictions |
regress |
Ordinary or weighted least-square regression |
Visualizing time series data
Function |
Description |
ascii |
Visualize one column of a time series as an ASCII chart |
plot |
Line chart for time series objects |