Added a forecasting recipe

[guilhermebodin]

#170 I think we could benefit from some simple plot recipes, no more searching how to make a proper ribbon.

Maybe we can even define more recipes for scenarios and components in the future.

Look what have I done in a feel lines

using StateSpaceModels, CSV, DataFrames, Plots
nile = CSV.read(StateSpaceModels.NILE, DataFrame)
model = LocalLevel(nile.flow)
fit!(model)
forec = forecast(model, 24)
plot(model, forec)

air_passengers = CSV.read(StateSpaceModels.AIR_PASSENGERS, DataFrame)
log_air_passengers = log.(air_passengers.passengers)
model = BasicStructural(log_air_passengers, 12)
fit!(model)
forec = forecast(model, 24)
plot(model, forec; legend = :topleft)

[codecov]

Codecov Report

Merging #223 (cb0ae70) into master (4f510fe) will increase coverage by 0.04%.
The diff coverage is 93.75%.

@@            Coverage Diff             @@
##           master     #223      +/-   ##
==========================================
+ Coverage   91.59%   91.63%   +0.04%     
==========================================
  Files          22       23       +1     
  Lines        1499     1531      +32     
==========================================
+ Hits         1373     1403      +30     
- Misses        126      128       +2     

Impacted Files	Coverage Δ
src/StateSpaceModels.jl	100.00% <ø> (ø)
src/visualization/forecast.jl	93.75% <93.75%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 4f510fe...cb0ae70. Read the comment docs.

[raphaelsaavedra]

I'm not familiar with RecipesBase, what's the difference between this and simply extending plot?

[guilhermebodin]

You don't need to have Plots.jl as a dependency.

[raphaelsaavedra]

Other than a small detail which I'm not sure is necessary, looks good to me!

[guilhermebodin]

I should test with this https://github.com/JuliaPhysics/Measurements.jl/blob/master/test/plots.jl