Lesson 2: AttributeError when trying to use plot_confusion_matrix

Part 1 2022
1

I’m following along with the fastai book. I encountered an error plotting the confusion matrix in chapter 2.

After training the learner for the first time I run the following code:

interp = ClassificationInterpretation.from_learner(learner)
learner.plot_confusion_matrix()

But I get the following error:

---------------------------------------------------------------------------

AttributeError                            Traceback (most recent call last)

<ipython-input-15-5ed97f736e90> in <cell line: 2>()
      1 interp = ClassificationInterpretation.from_learner(learner)
----> 2 learner.plot_confusion_matrix()

1 frames

/usr/local/lib/python3.9/dist-packages/torch/nn/modules/module.py in __getattr__(self, name)
   1612             if name in modules:
   1613                 return modules[name]
-> 1614         raise AttributeError("'{}' object has no attribute '{}'".format(
   1615             type(self).__name__, name))
   1616 

AttributeError: 'Sequential' object has no attribute 'plot_confusion_matrix'

You can see the complete notebook here:

Any ideas why this might be the case? I’m using Google Colab with the latest version of fastai (2.7.12).

2

[quote=“enigmeta, post:1, topic:105101”]

uential' object has no attribute 'plot_confusion_matrix'

Hi enigmeta,
you just confused the terms: the classificationinterpreter is (as you called) ‘interp’

So, just write:
interp.plot_confusion_matrix()

it should work :slight_smile:

2 Likes
3

Hello!

You need to use the Interpretation object in order to use its methods i.e. instead of learner.plot_confusion_matrix(), you need to do interp.plot_confusion_matrix().

I think this will resolve the error.

Hope this helps :slight_smile:

2 Likes
4

Thanks for the solution! I’ve been banging my head against the wall why it wouldn’t work. It helps to have somebody with fresh eyes to look at it.

Thanks again!

1 Like