I'm a little lost trying to figure out Prophet's diagnostics. Fitting model in a previous cell with interval_width=0.95 and a custom holidays dataframe that does not distinguish between holidays (just a flag like the playoffs df from the Peyton Manning example). Then running
from fbprophet.diagnostics import cross_validation
df_cv = cross_validation(model, initial = '365 days', period = '90 days', horizon = '30 days')
leads to the error
---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
<ipython-input-85-a50f3c0d9457> in <module>()
1 from fbprophet.diagnostics import cross_validation
----> 2 df_cv = cross_validation(model, initial = '270 days', period = '90 days', horizon = '30 days')
3 df_cv.head()
/usr/local/lib/python2.7/dist-packages/fbprophet/diagnostics.pyc in cross_validation(model, horizon, period, initial)
93 for cutoff in cutoffs:
94 # Generate new object with copying fitting options
---> 95 m = prophet_copy(model, cutoff)
96 # Train model
97 history_c = df[df['ds'] <= cutoff]
/usr/local/lib/python2.7/dist-packages/fbprophet/diagnostics.pyc in prophet_copy(m, cutoff)
139 raise Exception('This is for copying a fitted Prophet object.')
140
--> 141 if m.specified_changepoints:
142 changepoints = m.changepoints
143 if cutoff is not None:
AttributeError: 'Prophet' object has no attribute 'specified_changepoints'
I made sure the model exists, but a little confused on what to look for next. Thanks!
I'm a little lost trying to figure out Prophet's diagnostics. Fitting model in a previous cell with interval_width=0.95 and a custom holidays dataframe that does not distinguish between holidays (just a flag like the playoffs df from the Peyton Manning example). Then running
leads to the error
I made sure the model exists, but a little confused on what to look for next. Thanks!