Refreshing Data

[KristiGourlay]

Disclaimer: I'm a pythonista and my R is very basic.

I'm trying to refresh my model with 3 new weeks of data. I followed the demo and wrote:

RobynRefresh <- robyn_refresh(
  json_file = json_file,
  dt_input = refreshed_data,
  dt_holidays = dt_prophet_holidays,
  refresh_steps = 3, 
  refresh_iters = 1000, 
  refresh_trials = 3
)

Then I got an error for no object refreshStart detected and then for no object refreshEnd detected. So I updated to:

refreshStart <- '2023-01-08'
refreshEnd <- '2023-01-22'
units <- "weeks"

RobynRefresh <- robyn_refresh(
  json_file = json_file,
  dt_input = refreshed_data,
  dt_holidays = dt_prophet_holidays,
  refresh_steps = 3, 
  refresh_iters = 1000, 
  refresh_trials = 3
)

Now it runs. It successfully re-created my model 1_122_7. But then I get this:

>>> Building refresh model #1 in manual mode
>>> New bounds freedom: 100%
>> Running feature engineering...
NOTE: potential improvement on splitting channels for better exposure fitting. Threshold (Minimum R2) = 0.8 
  Check: InputCollect$plotNLSCollect outputs
  Weak relationship for: 'fbawareness_I', 'microsoft_I', 'mtv_I'and their spend
Input data has 264 weeks in total: 2018-01-07 to 2023-01-22
Refresh #1 model is built on rolling window of 3 week: 2018-01-28 to 2023-01-22
Rolling window moving forward: 3 weeks
Fitting time series with all available data...
Using geometric adstocking with 26 hyperparameters (25 to iterate + 1 fixed) on 9 cores
>>> Starting 3 trials with 1000 iterations each using TwoPointsDE nevergrad algorithm...
  Running trial 1 of 3
  |                                                                                   |   0%Timing stopped at: 0.531 0.52 0.32
Error in { : 
  task 1 failed - "arguments imply differing number of rows: 12, 13"

What I have tried:

• Confirmed no null values.
• Confirmed dimensions of data (261, 23) and now (264, 23)

[laresbernardo]

Creating an issue ticket. Please, let's follow the discussion over there.