You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I wanted to try your pyod wrapper. I am working with pyspark dataframes in databricks. the model_data is a spark dataframe with 24 floattype features (columns) and many rows. It is actually a time series (so this this wrapper also work for TODS ?) This is my code:
`from pyod.models.lof import LOF
from kakapo import PyodWrapper
from kakapo import get_default_model_space
I wanted to try your pyod wrapper. I am working with pyspark dataframes in databricks. the
model_data
is a spark dataframe with 24 floattype features (columns) and many rows. It is actually a time series (so this this wrapper also work for TODS ?) This is my code:`from pyod.models.lof import LOF
from kakapo import PyodWrapper
from kakapo import get_default_model_space
params = {
"type": "lof",
"n_neighbors":20,
"contamination": 0.01
}
model_space = {
"lof": LOF
}
model = PyodWrapper(**params)
model.set_model_space(model_space)
model.fit(model_data)`
When i simply try to fit the model on the dataframe i get the following error:
ValueError: Expected 2D array, got scalar array instead:
So my Question is, what type of dataframe should i pass into the model.fit() function / is there any more preprocessing necessary?
The text was updated successfully, but these errors were encountered: