Attributeerror: module 'numpy' has no attribute typedict (Solved)

Attributeerror: module ‘numpy’ has no attribute ‘typedict’ error occurs because ‘typedict’ is deprecated. We can use ‘sctypedict’ in the place of ‘typedict’ in the latest versions of numpy. In this article, we will explore the best ways to fix this error.

Attributeerror: module ‘numpy’ has no attribute typedict (Step by Step) –

Please follow the below steps to fix this error.

1. Use ‘sctypedict’ in the place of ‘typedict’ –

Since the release notes of Numpy 1.21.0 clearly states that ‘typedict’ is not supportive in the versions above to it. Hence it is advisable to replace np.typeDict by np.sctypeDict in the codebase we have.

Attributeerror module 'numpy' has no attribute typedict solved by relplacing sctypedict — Attributeerror module ‘numpy’ has no attribute typedict solved by relplacing sctypedict

2. Downgrade Numpy Version to lesser or equal 1.21.0 version –

In general, this type of error occurs mainly because of code depreciation or code structure change in the module. Then downgrading is the best and easiest way to fix it. It will also help here as well. In most of the code communities, multiple developers confirmed that downgrading really helped them in resolving this error. We can use pip for downgrading numpy. Please refer the below command.

pip install numpy==1.21

It will downgrade numpy to 1.21 version.

3. If numpy is not directly involved –

In some cases it happens that we are not using numpy module directly but the other Library we use internally involve numpy library. Now that end library is pointing to certain numpy version which is causing incompatibility. In this type of situation, we upgrade the the end library which is invoking numpy. After upgrade the end library will point different version of numpy and it will resolve the error too.

For example we are getting this error in h5 module then we can upgrade the h5 module like below.