**I can not understand what I have to do.**

This error means that your implementation of calculating the information gain is not right, what you have to do is go back and redo the function or exerciseâ€¦

First of all if I did any mistake then how expected example satisfy with my result. However here is my code that I have written. Please take a look and help me to figure out the solution.

```
# UNQ_C3
# GRADED FUNCTION: compute_information_gain
def compute_information_gain(X, y, node_indices, feature):
"""
Compute the information of splitting the node on a given feature
Args:
X (ndarray): Data matrix of shape(n_samples, n_features)
y (array like): list or ndarray with n_samples containing the target variable
node_indices (ndarray): List containing the active indices. I.e, the samples being considered in this step.
Returns:
cost (float): Cost computed
"""
# Split dataset
left_indices, right_indices = split_dataset(X, node_indices, feature)
# print("left_indices:", left_indices, "right_indices:" , right_indices, "X ->", X, "feature->", feature)
# Some useful variables
X_node, y_node = X[node_indices], y[node_indices]
X_left, y_left = X[left_indices], y[left_indices]
X_right, y_right = X[right_indices], y[right_indices]
# You need to return the following variables correctly
information_gain = 0
### START CODE HERE ###
# Moderator edit: code removed
### END CODE HERE ###
return information_gain
```

Please donâ€™t post your code on the forum. The Code of Conduct does not allow it.

If a mentor needs to see your code, weâ€™ll contact you with instructions.

The other possibility is maybe your compute_entropy() or split_dataset() code doesnâ€™t work correctly.

Passing the test cases built into the notebook does not prove your code is perfect.

The data set used by compute_information_gain_test() is different. So that may expose defects in your other functions.

i have the same issue if it have been resolved for you how can i resolve it

I had faced similar problem. The issue is that compute_entropy function needs to cover all scenarios such as if the dataset is completely pure etc.

The compute_information_gain function uses compute_entropy, and if compute_entropy doesnâ€™t have all edge cases covered, then compute_information_gain will also fail.

Regards

Nagesh

Canâ€™t paste my code here :).