'numpy.float32' object has no attribute 'is_integer' about chefboost

[serengil]

could you share your data set?

[Gabomfim]

Sorry for keeping you waiting.

I'm sharing with you my notebook with all the files, including the databases used (in the data file).
I managed to fix the problem by importing the database as txt instead of csv.

allstroke.txt is the txt version of the healthcare-dataset-stroke-data.csv database. That did the fix.

We now import the database in this way:
df = pd.read_csv("./data/allStroke.txt", index_col=0)

I don't have the old code with me now, but I can send it to you the next week if needed.

[serengil]

When I run this in my environment, it works well. I have Python 3.8.12, pandas==1.3.5. I recommend you to upgrade or downgrade to my environment level.

from chefboost import Chefboost as chef
import pandas as pd

df = pd.read_csv("healthcare-dataset-stroke-data.csv", index_col=0)

print(df.head())

configGBM = {'algorithm': 'C4.5', 'enableGBM': True, 'epochs': 7, 'learning_rate': 1, 'max_depth': 5, 'enableParallelism': False}
modelGBM = chef.fit(df = df, config = configGBM)

Output logs:
(sefik) sefik@Sefiks-MacBook-Pro Desktop % python hello.py
gender age hypertension heart_disease ever_married work_type Residence_type avg_glucose_level bmi smoking_status Decision
id
9046 Male 67.0 0 1 Yes Private Urban 228.69 36.6 formerly smoked Yes
51676 Female 61.0 0 0 Yes Self-employed Rural 202.21 NaN never smoked Yes
31112 Male 80.0 0 1 Yes Private Rural 105.92 32.5 never smoked Yes
60182 Female 49.0 0 0 Yes Private Urban 171.23 34.4 smokes Yes
1665 Female 79.0 1 0 Yes Self-employed Rural 174.12 24.0 never smoked Yes
Gradient Boosting Machines...
Regression tree is going to be built...
gradient boosting for classification
Epoch 7. Accuracy: 82. Process: : 100%|█████████████████████████████████████████████████████████████████████████████████████| 7/7 [00:42<00:00, 6.02s/it]
The best accuracy got in 6 epoch with the score 82.78210116731518

finished in 42.12960386276245 seconds

Evaluate train set

Accuracy: 82.00389105058366 % on 1028 instances
Labels: ['Yes' 'No']
Confusion matrix: [[99, 35], [150, 744]]
Precision: 73.8806 %, Recall: 39.759 %, F1: 51.6971 %