Hey together,
I am replicating some parts of a paper in the GaR literature. I successfully ran the quantile regressions and stored the predicted values in each new variable, respectively (each variable reflects a certain quantile y_q over time). The problem is that the data on the y-axis range both in positive and negative values (also, there are some quantile, which have both positive and negative values). This is the code
which delivers
When I add the "nodropbase"-command to the code, I receive the following:
This still looks not like the graph I try to replicate. I am fully aware that it must have something to do with the base-command. I tried to follow https://www.statalist.org/forums/for...-cyclic-series, but I am not sure if this is the right procedure.
Can anyone help me?
Best,
Mario
I am replicating some parts of a paper in the GaR literature. I successfully ran the quantile regressions and stored the predicted values in each new variable, respectively (each variable reflects a certain quantile y_q over time). The problem is that the data on the y-axis range both in positive and negative values (also, there are some quantile, which have both positive and negative values). This is the code
Code:
twoway area y_q5 y_q10 y_q25 yq50 y_q75 y_q90 y_q95 date, xtitle("") xlabel(,format(%tdCY)) title("One quarter ahead")
When I add the "nodropbase"-command to the code, I receive the following:
This still looks not like the graph I try to replicate. I am fully aware that it must have something to do with the base-command. I tried to follow https://www.statalist.org/forums/for...-cyclic-series, but I am not sure if this is the right procedure.
Can anyone help me?
Best,
Mario
Comment