Hello All!

I am working with sectoral employment data pulled from the FRED using "import fred". After initially creating a time series plot with the raw data, I decided graphing sectoral employment as a year to date change would be a more helpful visual. When I lag my variables, the chart axis shift completely (I've attached a picture of the graph error - data set is from 2016-current).


Could someone tell me where my error is? All variables are collected monthly. The weirdest thing is that when I rerun the code with a different daterange chosen when importing data the axis changes how shifted it is (two pictures one created with the import daterange starting at 2016 and one at 2018.

************************************************** ************************************************** ************************************************** ************************************************** ********
import fred v1 v 2 v3, daterange(2016-01-01 .)
replace daten = mofd(daten)
gen lag = daten[_n-12]
tsset lag
***************Percent Change Lag**********
generate v1_lag = ((v1[_n] - v1[_n-12])/ v1[_n-12])*100
generate v2_lag = ((v2[_n] - v2[_n-12])/ v2[_n-12])*100
generate v3_lag = ((v3[_n] - v3[_n-12])/ v3[_n-12])*100
graph twoway (tsline v1_lag) (tsline v2_lag) (tsline v3_lag), ytitle("Percent Change from Last Year" "") xtitle("" "") title("Percent Change in Number of People Employed by Sector") subtitle("January 2016-June 2020") note("Source: Federal Reserve Bank of St Louis ") tlabel(, format(%tmCCYY))]

************************************************** ************************************************** ************************************************** ************************************************** ********

[ATTACH]temp_18885_1594309031634_989[/ATTACH]

I followed this code but for ease of reading I renamed the variables - removed one variable from sample code since it is occurring in all of them on the graph