As stated above, I'm attempting to do ordinal logistic regression that is attempting to explain a rating (rating scale from 4 to 9) using time series data. Here is a snippet of my code:
from mord import LogisticAT
import numpy as np
import pandas as pd
from pandas import DataFrame
df = pd.read_csv("mydata.csv",error_bad_lines=False, index_col=False,
dtype='unicode',low_memory=False)
y = list(map(int,np.array(df.rating)))
x = np.transpose([np.array(df['series1']),np.array(df['series2'])])
model = LogisticAT(alpha=0)
model.fit(x,y)
When I do this I get the following error:
"y_tmp = y - y.min() # we need classes that start at zero
AttributeError: 'list' object has no attribute 'min'"
Now I made sure I scaled my ratings such that they start at 0 and go to up to 5 and it still didn't help. I'm thinking the issue is caused by the second part, but my dependept variable clearly has a min.
question from:
https://stackoverflow.com/questions/65848689/having-issues-with-logisticat-modeling-from-mord-library 与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…