+ Added `scatter.hist` to draw a scatterplot with marginal histograms - never really liked other versions.
+ Minor updates to `tsplot`, `SigExtract` and `LagReg` to improve the displays.
+ Added scripts for the analysis of DNA sequences and other categorical time series: `specenv`, `dna2vector`, and the data set `EBV`. And `specenv` can also handle real-valued series.
- `dna2vector` is used to preprocess a categorical sequence.
- `EBV` is the entire Epstein-Barr sequence as a long single string. It's not useful on its own, but thru `dna2vector`, different regions can be explored via `specenv`.
+ We needed powers of matrices enough where we thought we'd include it in `astsa`. The script is called `matrixpwr` and includes `%^%` as a more intuitive operator. For example, `var(econ5)%^%-.5` to calculate an inverse square root matrix.
+ Added `test.linear`, a script to test the null hypothesis that the data are generated from a linear process with iid innovations.
+ Updated `Grid`, `tsplot`, `sarima.sim`, and `mvspec`.
- `Grid` and `tsplot` will produce grid lines at the minor ticks. These can be shut off individually on either axis.
- For `sarima.sim`, now allow seasonal period without having to specify
other seasonal parameters - doing so gives a message to make sure you're doing it on purpose, whereas it used to stop the execution. There's an example of the advantage of this in its man page.
- For `tsplot` and `mvspec`, by default now, there is one minor tick on the x-axis and none on the y-axis. Also, `mvspec` doesn't display the bandwidth
on the axis - it's still there in the CI if the plot is on log-scale and it's still part of the "spec" object.
- Also, updated `tsplot` so multiple series can have different plot symbols (`pch`), e.g., `tsplot(blood, type='o', col=2:4, pch=2:4, cex=2)`
+ Prettified `arma.spec`, `lag1.plot`, `lag2.plot`, `sarima`, and `sarima.for`, using the awesome power of `tsplot`, but no need to change existing code.
+ Updated `tsplot` to allow for spaghetti plots:<br/>
`x <- replicate(100, cumsum(rcauchy(1000))/1:1000)`<br/>
`tsplot(x, col=1:8, spaghetti=TRUE)`
+ The package now has its own color palette that is attached
when the package is attached. The palette is especially suited for plotting
time series. It is a bit darker than the new default R4 palette.
You can revert back using `palette("default")`.
- In addition, added `astsa.col` script to easily adjust opacity of the astsa color palette - examples on its man page.
+ Improvements to `acf1`, `acf2`, and `ccf2` employing the power of
`tsplot` to give some kickass displays. For example,
`acf2(soi, col=4, lwd=3, gg=TRUE)`.
+ `sarima.sim` output used to start at time `0` - now the start
time is up to the user (with default `t0=0`).