You can use `#fmt: on/off` as explained in the issue linked. In your case it would look like:

```python
# fmt: off
np.array(
    [
        [1, 0, 0, 0],
        [0, -1, 0, 0],
        [0, 0, 1, 0],
        [0, 0, 0, -1],
    ]
)
# fmt: on
```

`# fmt: off` disables formatting for all following lines until formatting is activated again with `# fmt: on`

If you&#39;re willing to change your code slightly, then Black leaves either of the following alone:

    contents = [
        [1, 0, 0, 0],
        [0, -1, 0, 0],
        [0, 0, 1, 0],
        [0, 0, 0, -1],
    ]

    np.array(contents)

This is because the trailing comma in the multi-line list is magic. [Black takes it to mean that you plan to extend the list in future](https://black.readthedocs.io/en/stable/the_black_code_style/current_style.html#the-magic-trailing-comma), although in this case it just means Black&#39;s style isn&#39;t very readable. Unfortunately the trailing comma isn&#39;t magic *enough* to work when the list is wrapped in that extra function call.

    np.array(
        [
            # just say anything
            [1, 0, 0, 0],
            [0, -1, 0, 0],
            [0, 0, 1, 0],
            [0, 0, 0, -1],
        ]
    )

This is because Black cannot outwit Python&#39;s lack of inline comments!

The latest version of black ( &gt;= 21.0) takes into account the comma after the last element.

#### So:
```python
np.array(
    [
        [1, 0, 0, 0],
        [0, -1, 0, 0],
        [0, 0, 1, 0],
        [0, 0, 0, -1]
    ]
)
```
will be formatted to:
```python
np.array([[1, 0, 0, 0], [0, -1, 0, 0], [0, 0, 1, 0], [0, 0, 0, -1]])
```
(note no last comma)

#### Instead
 
```python
np.array([[1, 0, 0, 0], [0, -1, 0, 0], [0, 0, 1, 0], [0, 0, 0, -1],])
```
will be formatted to:
```python
np.array(
    [
        [1, 0, 0, 0],
        [0, -1, 0, 0],
        [0, 0, 1, 0],
        [0, 0, 0, -1],
    ]
)
```
(note last comma)

I&#39;m having a hard time understanding the &#39;exhaustive-deps&#39; lint rule.

I already read [this post][1] and [this post][2] but I could not find an answer.

Here is a simple React component with the lint issue: 

    const MyCustomComponent = ({onChange}) =&gt; {
        const [value, setValue] = useState(&#39;&#39;);
    
        useEffect(() =&gt; {
            onChange(value);
        }, [value]);
    
        return (
            &lt;input 
               value={value} 
               type=&#39;text&#39; 
               onChange={(event) =&gt; setValue(event.target.value)}&gt;
            &lt;/input&gt;
        )
    } 
    
It requires me to add `onChange` to the `useEffect` dependencies array. But in my understanding `onChange` will never change, so it should not be there.

Usually I manage it like this: 

    const MyCustomComponent = ({onChange}) =&gt; {
        const [value, setValue] = useState(&#39;&#39;);
    
        const handleChange = (event) =&gt; {
            setValue(event.target.value);
            onChange(event.target.value)
        }
    
        return (
            &lt;input 
               value={value} 
               type=&#39;text&#39;
               onChange={handleChange}&gt;
            &lt;/input&gt; ​
        )
    } 

Why the lint? Any clear explanation about the lint rule for the first example?

*Or should I not be using `useEffect` here? (I&#39;m a noob with hooks)*

  [1]: https://stackoverflow.com/questions/58549846/react-useeffect-hook-with-warning-react-hooks-exhaustive-deps
  [2]: https://stackoverflow.com/questions/57983717/designing-react-hooks-prevent-react-hooks-exhaustive-deps-warning


Understanding the React Hooks &#39;exhaustive-deps&#39; lint rule

How can I use a React ref as a mutable instance, with Typescript? The current property appears to be typed as read-only.

I am using React + Typescript to develop a library that interacts with input fields that are NOT rendered by React. I want to capture a reference to the HTML element and then bind React events to it. 


```tsx
  const inputRef = useRef&lt;HTMLInputElement&gt;();
  const { elementId, handler } = props;

  // Bind change handler on mount/ unmount
  useEffect(() =&gt; {
    inputRef.current = document.getElementById(elementId);
    if (inputRef.current === null) {
      throw new Exception(`Input with ID attribute ${elementId} not found`);
    }
    handler(inputRef.current.value);

    const callback = debounce((e) =&gt; {
      eventHandler(e, handler);
    }, 200);

    inputRef.current.addEventListener(&#39;keypress&#39;, callback, true);

    return () =&gt; {
      inputRef.current.removeEventListener(&#39;keypress&#39;, callback, true);
    };
  });
```

It generates compiler errors: `semantic error TS2540: Cannot assign to &#39;current&#39; because it is a read-only property.`

I also tried `const inputRef = useRef&lt;{ current: HTMLInputElement }&gt;();` This lead to this compiler error:

```
Type &#39;HTMLElement | null&#39; is not assignable to type &#39;{ current: HTMLInputElement; } | undefined&#39;.

  Type &#39;null&#39; is not assignable to type &#39;{ current: HTMLInputElement; } | undefined&#39;.
```

What typescript type do I use with useRef() hook when setting current manually?

I would like to ignore a specific multi-line code by `black` python formatter. Particularly, this is used for `np.array` or matrix construction which turned ugly when formatted. Below is the example.

```
np.array(
    [
        [1, 0, 0, 0],
        [0, -1, 0, 0],
        [0, 0, 1, 0],
        [0, 0, 0, -1],
    ]
)
# Will be formatted to
np.array([[1, 0, 0, 0], [0, -1, 0, 0], [0, 0, 1, 0], [0, 0, 0, -1]])
```

I found [this](https://github.com/psf/black/issues/451) issue in `black` github, but that only works for inline command, which is not what I have here.

Is there anything I can do to achieve this for a multi-line code?

Black formatter - Ignore specific multi-line code

I would like to ignore a specific multi-line code by <code>black</code> python formatter. Particularly, this is used for <code>np.array</code> or matrix construction which turned ugly when formatted. Below is the example.
<pre><code class="hljs language-css">np.array(
 [ [1, 0, 0, 0],
 [0, -1, 0, 0],
 [0, 0, 1, 0],
 [0, 0, 0, -1],
 ]
)
# Will be formatted to
np.array([[1, 0, 0, 0], [0, -1, 0, 0], [0, 0, 1, 0], [0, 0, 0, -1]])
</code></pre>
I found <a href="https://github.com/psf/black/issues/451" target="_blank" rel="noopener noreferrer">this</a> issue in <code>black</code> github, but that only works for inline command, which is not what I have here.
Is there anything I can do to achieve this for a multi-line code?

I&#39;m setting up an autoclicker in Python 3.8 and I need win32api for GetAsyncKeyState but it always gives me this error:

```
&gt;&gt;&gt; import win32api
Traceback (most recent call last):
  File &quot;&lt;stdin&gt;&quot;, line 1, in &lt;module&gt;
ImportError: DLL load failed while importing win32api: The specified module could not be found.
```

I&#39;m on Windows 10 Home 64x. I&#39;ve already tried

```
pip install pypiwin32
```

And it successfully installs but nothing changes. I tried uninstalling and re-installing python as well. I also tried installing &#39;django&#39; in the same way and it actually works when I `import django`, so I think it&#39;s a win32api issue only.


```
&gt;&gt;&gt; import win32api
```

I expect the output to be none, but the actual output is always that error ^^

How to fix &quot;ImportError: DLL load failed&quot; while importing win32api

Continuation from previous question: https://stackoverflow.com/questions/58635521/tensorflow-typeerror-int-object-is-not-iterable/58635565#58635583

My training data is a list of lists each comprised of 1000 floats. For example, `x_train[0] =`

    [0.0, 0.0, 0.1, 0.25, 0.5, ...]

Here is my model:

    model = Sequential()
    
    model.add(LSTM(128, activation=&#39;relu&#39;,
                   input_shape=(1000, 1), return_sequences=True))
    model.add(Dropout(0.2))
    model.add(LSTM(128, activation=&#39;relu&#39;))
    model.add(Dropout(0.2))
    model.add(Dense(32, activation=&#39;relu&#39;))
    model.add(Dropout(0.2))
    model.add(Dense(1, activation=&#39;sigmoid&#39;))
    
    opt = tf.keras.optimizers.Adam(lr=1e-3, decay=1e-5)
    
    model.compile(optimizer=&#39;rmsprop&#39;,
                  loss=&#39;binary_crossentropy&#39;,
                  metrics=[&#39;accuracy&#39;])
    
    model.fit(x_train, y_train, epochs=3, validation_data=(x_test, y_test))

 Here is the error I&#39;m getting:
      
  

    Traceback (most recent call last):
          File &quot;C:\Users\bencu\Desktop\ProjectFiles\Code\Program.py&quot;, line 88, in FitModel
            model.fit(x_train, y_train, epochs=3, validation_data=(x_test, y_test))
          File &quot;C:\Users\bencu\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow_core\python\keras\engine\training.py&quot;, line 728, in fit
            use_multiprocessing=use_multiprocessing)
          File &quot;C:\Users\bencu\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow_core\python\keras\engine\training_v2.py&quot;, line 224, in fit
            distribution_strategy=strategy)
          File &quot;C:\Users\bencu\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow_core\python\keras\engine\training_v2.py&quot;, line 547, in _process_training_inputs
            use_multiprocessing=use_multiprocessing)
          File &quot;C:\Users\bencu\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow_core\python\keras\engine\training_v2.py&quot;, line 606, in _process_inputs
            use_multiprocessing=use_multiprocessing)
          File &quot;C:\Users\bencu\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow_core\python\keras\engine\data_adapter.py&quot;, line 479, in __init__
            batch_size=batch_size, shuffle=shuffle, **kwargs)
          File &quot;C:\Users\bencu\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow_core\python\keras\engine\data_adapter.py&quot;, line 321, in __init__
            dataset_ops.DatasetV2.from_tensors(inputs).repeat()
          File &quot;C:\Users\bencu\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow_core\python\data\ops\dataset_ops.py&quot;, line 414, in from_tensors
            return TensorDataset(tensors)
          File &quot;C:\Users\bencu\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow_core\python\data\ops\dataset_ops.py&quot;, line 2335, in __init__
            element = structure.normalize_element(element)
          File &quot;C:\Users\bencu\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow_core\python\data\util\structure.py&quot;, line 111, in normalize_element
            ops.convert_to_tensor(t, name=&quot;component_%d&quot; % i))
          File &quot;C:\Users\bencu\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow_core\python\framework\ops.py&quot;, line 1184, in convert_to_tensor
            return convert_to_tensor_v2(value, dtype, preferred_dtype, name)
          File &quot;C:\Users\bencu\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow_core\python\framework\ops.py&quot;, line 1242, in convert_to_tensor_v2
            as_ref=False)
          File &quot;C:\Users\bencu\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow_core\python\framework\ops.py&quot;, line 1296, in internal_convert_to_tensor
            ret = conversion_func(value, dtype=dtype, name=name, as_ref=as_ref)
          File &quot;C:\Users\bencu\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow_core\python\framework\tensor_conversion_registry.py&quot;, line 52, in _default_conversion_function
            return constant_op.constant(value, dtype, name=name)
          File &quot;C:\Users\bencu\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow_core\python\framework\constant_op.py&quot;, line 227, in constant
            allow_broadcast=True)
          File &quot;C:\Users\bencu\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow_core\python\framework\constant_op.py&quot;, line 235, in _constant_impl
            t = convert_to_eager_tensor(value, ctx, dtype)
          File &quot;C:\Users\bencu\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow_core\python\framework\constant_op.py&quot;, line 96, in convert_to_eager_tensor
            return ops.EagerTensor(value, ctx.device_name, dtype)
        ValueError: Failed to convert a NumPy array to a Tensor (Unsupported object type float).

I&#39;ve tried googling the error myself, I found something about using the `tf.convert_to_tensor` function. I tried passing my training and testing lists through this but the function won&#39;t take them.

Tensorflow - ValueError: Failed to convert a NumPy array to a Tensor (Unsupported object type float)

Today when I tried to run a simple code on Sublime Text 3, the following message appeared:

&gt; Python was not found but can be installed from the Microsoft Store: https://go.microsoft.com/fwlink?linkID=2082640

And when I type Python in CMD, it opens the Windows Store for me to download Python 3.7. This problem started today for no good reason, I didn&#39;t change or downloaded anything about Python and already tried reinstalling Python, and the Path is correct.

CMD opens window store when I type python

[`TypeVar`](https://docs.python.org/3/library/typing.html#typing.TypeVar) and [`NewType`](https://docs.python.org/3/library/typing.html#newtype) seem related but I&#39;m not sure when I&#39;m supposed to use each or what the difference is at runtime and statically.

What is the difference between TypeVar and NewType?

I am trying to train the following CNN as follows, but I keep getting the same error regarding .cuda() and I am not sure how to fix it. Here is a chunk of my code so far.

```
import matplotlib.pyplot as plt
import numpy as np
import torch
from torch import nn
from torch import optim
import torch.nn.functional as F
import torchvision
from torchvision import datasets, transforms, models
from torch.utils.data.sampler import SubsetRandomSampler


data_dir = &quot;/home/ubuntu/ML2/ExamII/train2/&quot;
valid_size = .2

# Normalize the test and train sets with torchvision
train_transforms = transforms.Compose([transforms.Resize(224),
                                           transforms.ToTensor(),
                                           ])

test_transforms = transforms.Compose([transforms.Resize(224),
                                          transforms.ToTensor(),
                                          ])

# ImageFolder class to load the train and test images
train_data = datasets.ImageFolder(data_dir, transform=train_transforms)
test_data = datasets.ImageFolder(data_dir, transform=test_transforms)


# Number of train images
num_train = len(train_data)
indices = list(range(num_train))
# Split = 20% of train images
split = int(np.floor(valid_size * num_train))
# Shuffle indices of train images
np.random.shuffle(indices)
# Subset indices for test and train
train_idx, test_idx = indices[split:], indices[:split]
# Samples elements randomly from a given list of indices
train_sampler = SubsetRandomSampler(train_idx)
test_sampler = SubsetRandomSampler(test_idx)
# Batch and load the images
trainloader = torch.utils.data.DataLoader(train_data, sampler=train_sampler, batch_size=1)
testloader = torch.utils.data.DataLoader(test_data, sampler=test_sampler, batch_size=1)


#print(trainloader.dataset.classes)

device = torch.device(&quot;cuda&quot; if torch.cuda.is_available() else &quot;cpu&quot;)
model = models.resnet50(pretrained=True)

model.fc = nn.Sequential(nn.Linear(2048, 512),
                                 nn.ReLU(),
                                 nn.Dropout(0.2),
                                 nn.Linear(512, 10),
                                 nn.LogSigmoid())
                                 # nn.LogSoftmax(dim=1))
# criterion = nn.NLLLoss()
criterion = nn.BCELoss()
optimizer = optim.Adam(model.fc.parameters(), lr=0.003)
model.to(device)

#Train the network
for epoch in range(2):  # loop over the dataset multiple times

    running_loss = 0.0
    for i, data in enumerate(trainloader, 0):
        # get the inputs; data is a list of [inputs, labels]
        inputs, labels = data

        # zero the parameter gradients
        optimizer.zero_grad()

        # forward + backward + optimize
        outputs = model(inputs)
        loss = criterion(outputs, labels)
        loss.backward()
        optimizer.step()

        # print statistics
        running_loss += loss.item()
        if i % 2000 == 1999:    # print every 2000 mini-batches
            print(&#39;[%d, %5d] loss: %.3f&#39; %
                  (epoch + 1, i + 1, running_loss / 2000))
            running_loss = 0.0

print(&#39;Finished Training&#39;)
```

However, I keep getting this error in the console: 
&gt;RuntimeError: Input type (torch.FloatTensor) and weight type (torch.cuda.FloatTensor) should be the same`

Any thoughts on how to fix it? I read that maybe the model hasn&#39;t been pushed into my GPU, but not sure how to fix it. Thanks!

Content Type	Original Author	Original Content on Stackoverflow
Question	Darren Christopher	View Question on Stackoverflow
Solution 1 - Python	OriolAbril	View Answer on Stackoverflow
Solution 2 - Python	Steve Jessop	View Answer on Stackoverflow
Solution 3 - Python	SeF	View Answer on Stackoverflow

Black formatter - Ignore specific multi-line code

Python Problem Overview

Python Solutions

Solution 1 - Python

Solution 2 - Python

Solution 3 - Python

So:

Instead

What typescript type do I use with useRef() hook when setting current manually?

Understanding the React Hooks 'exhaustive-deps' lint rule

Attributions