Hi.

Is there any plans to license this project?

Thanks for your implementation！
When I want to train my model，I have a x_train(32300,92160) y_train(32300,1),but I got an error:

- Scanning and fitting for X ((32300,)) and y ((32300,)) started
- Window shape: [1, 23040] Total windows: 69121
Traceback (most recent call last):
File "gcForest_fer.py", line 71, in
mgc_forest.fit(x_train, y_train)
File "/home/wangjian/wj/deep_forest.py", line 68, in fit
for mgs in self.mgs_instances
File "/home/wangjian/wj/deep_forest.py", line 191, in scan
sliced_X, sliced_y = self.slices(X, y)
File "/home/wangjian/wj/deep_forest.py", line 154, in slices
windows_slices_list = X[(slice(None),) + axis_slices]
IndexError: too many indices for array.
Can you tell me why I am wrong? Thank you!

Hi, thank you for your implementation.

Based on my understanding of the paper, the length of the vector to each level in figure 1 of the paper should be the same. According to the paper "The estimated class distribution forms a class vector, which is then concatenated with the original feature vector to be input to the next level of cascade".

However, in your implementation, the input to the first level is 1520, the input to the second level is 1560, the input to the third level is 1600, etc.

If I misunderstood the paper, please kindly tell me.

Anyway, I fixed the code myself and find no significant difference in accuracy.

Shouldn't it be 11?

Thaks for your implementation!

When I want to train full of datasets as MNIST, example code is crashed because out of memory.
I commented out some lines that are limit the size of dataset about following codes.

#
# Limit the size of the dataset
#
# X_train = X_train[:2000]
# y_train = y_train[:2000]
# X_test = X_test[:2000]
# y_test = y_test[:2000]

Is these changes something to wrong?
Or does my computer have insufficient memory?

My testing computer is following specs.
OS: Windows 7
CPU: Core i7 970
Memory: 32GB (I only use real memories.)

If this problem causes insufficient memory, I want to know how to economize memory.
(e.g. like mini-batch training in deep neural network)

Code: http://lamda.nju.edu.cn/code_gcForest.ashx
Discussion: https://www.reddit.com/r/MachineLearning/comments/6ewdz7/r_deep_forest_towards_an_alternative_to_deep/

Code is hugged to death at the moment unfortunately.

Thank you for your completion！
err：
A worker process managed by the executor was unexpectedly terminated. This could be caused by a segmentation fault while calling the function or by an excessive memory usage causing the Operating System to kill the worker. The exit codes of the workers are {SIGKILL(-9)}
My computer's memory is 32g.
Have you ever encountered this situation?
Looking forward to your reply！

I am confused to the training and testing architectures of SDF and the implementation details of the experiments. I am long to figure the detail out in order to do further research in this field with your team's method. looking forward your reply

Hi,

following the tutorial in deep-forests-example.ipynb, at the end I tried to save the model using something like with open('out.pkl', 'wb') as pickle_file: pickle.dump(mgc_forest, pickle_file, protocol=2) , but then I get TypeError: can't pickle _thread.RLock objects.

Is there a way to save the model for later use?

Hi,
currently, I am trying use this deep forest model to test the EEG signal for the sleep stage classification. However, the dataset is 1-d, not the 2-d like image. So, could you lend me a hand, how can I change the code and make it possible for 1-d datasets training.
Great thanks!!!