Time series second order differences

timeseries
pandas
time_series
python

#1

Hi
So in this tutorial(https://www.analyticsvidhya.com/blog/2016/02/time-series-forecasting-codes-python/) it was written that in order to make sure even more that the series is stationary then we can also take second or third order difference.

So how can we do that?
I’ve tried applying the formula for 2nd order difference:
ts_log_diff = ts_log - 2*ts_log.shift()+ ts_log.shift(periods=2)

but both critical values and test statistics ended extremely small:
Results of Dickey-Fuller Test:
Test Statistic -8.196629e+00
p-value 7.419305e-13
Lags Used 1.300000e+01
Number of Observations Used 1.280000e+02
Critical Value (10%) -2.578960e+00
Critical Value (1%) -3.482501e+00
Critical Value (5%) -2.884398e+00

Is this done right?