Boa很适合您的需求，请参阅扩展Python tensorflow keras的示例。JavaScript中的顺序类。

const fs = require('fs');
const boa = require('@pipcook/boa');
const { tuple, enumerate } = boa.builtins();

const tf = boa.import('tensorflow');
const tfds = boa.import('tensorflow_datasets');

const { keras } = tf;
const { layers } = keras;

const [
  [ train_data, test_data ],
  info
] = tfds.load('imdb_reviews/subwords8k', boa.kwargs({
  split: tuple([ tfds.Split.TRAIN, tfds.Split.TEST ]),
  with_info: true,
  as_supervised: true
}));

const encoder = info.features['text'].encoder;
const padded_shapes = tuple([
  [ null ], tuple([])
]);
const train_batches = train_data.shuffle(1000)
  .padded_batch(10, boa.kwargs({ padded_shapes }));
const test_batches = test_data.shuffle(1000)
  .padded_batch(10, boa.kwargs({ padded_shapes }));

const embedding_dim = 16;
const model = keras.Sequential([
  layers.Embedding(encoder.vocab_size, embedding_dim),
  layers.GlobalAveragePooling1D(),
  layers.Dense(16, boa.kwargs({ activation: 'relu' })),
  layers.Dense(1, boa.kwargs({ activation: 'sigmoid' }))
]);

model.summary();
model.compile(boa.kwargs({
  optimizer: 'adam',
  loss: 'binary_crossentropy',
  metrics: [ 'accuracy' ]
}));

完整的示例在:https://github.com/alibaba/pipcook/blob/master/example/boa/tf2/word-embedding.js

我在另一个项目pipook中使用了Boa，这是为了解决JavaScript开发人员的机器学习问题，我们通过Boa库在Python生态系统(tensorflow,keras,pytorch)上实现了ML/DL模型。

你可以把你的python编译，然后像调用javascript一样调用它。我已经成功地为screeps做了这件事，甚至让它在浏览器中运行la brython。

之前的大多数答案都将承诺的成功称为on(“数据”)，这不是正确的方法，因为如果你收到很多数据，你只会得到第一部分。相反，你必须在end事件上做。

const { spawn } = require('child_process');
const pythonDir = (__dirname + "/../pythonCode/"); // Path of python script folder
const python = pythonDir + "pythonEnv/bin/python"; // Path of the Python interpreter

/** remove warning that you don't care about */
function cleanWarning(error) {
    return error.replace(/Detector is not able to detect the language reliably.\n/g,"");
}

function callPython(scriptName, args) {
    return new Promise(function(success, reject) {
        const script = pythonDir + scriptName;
        const pyArgs = [script, JSON.stringify(args) ]
        const pyprog = spawn(python, pyArgs );
        let result = "";
        let resultError = "";
        pyprog.stdout.on('data', function(data) {
            result += data.toString();
        });

        pyprog.stderr.on('data', (data) => {
            resultError += cleanWarning(data.toString());
        });

        pyprog.stdout.on("end", function(){
            if(resultError == "") {
                success(JSON.parse(result));
            }else{
                console.error(`Python error, you can reproduce the error with: \n${python} ${script} ${pyArgs.join(" ")}`);
                const error = new Error(resultError);
                console.error(error);
                reject(resultError);
            }
        })
   });
}
module.exports.callPython = callPython;

电话:

const pythonCaller = require("../core/pythonCaller");
const result = await pythonCaller.callPython("preprocessorSentiment.py", {"thekeyYouwant": value});

python:

try:
    argu = json.loads(sys.argv[1])
except:
    raise Exception("error while loading argument")

我知道的最简单的方法是使用“child_process”包，它随node一起打包。

然后你可以这样做:

const spawn = require("child_process").spawn;
const pythonProcess = spawn('python',["path/to/script.py", arg1, arg2, ...]);

然后你要做的就是确保你在python脚本中导入了sys，然后你就可以使用sys访问arg1了。Argv [1]， arg2使用sys。Argv[2]，等等。

要将数据发送回节点，只需在python脚本中执行以下操作:

print(dataToSendBack)
sys.stdout.flush()

然后node可以使用以下命令监听数据:

pythonProcess.stdout.on('data', (data) => {
 // Do something with the data returned from python script
});

由于这允许使用spawn将多个参数传递给脚本，您可以重新构造python脚本，以便其中一个参数决定调用哪个函数，而另一个参数传递给该函数，等等。

希望这是清楚的。如果有需要澄清的地方请告诉我。

许多例子都是过时的，并且涉及复杂的设置。您可以尝试JSPyBridge/pythonia(完全披露:我是作者)。它是一种普通的JS，可以让你操作外部Python对象，就好像它们存在于JS中一样。事实上，它实现了互操作性，因此Python代码可以通过回调和传递函数返回调用JS。

numpy + matplotlib的例子，用ES6导入系统:

import { py, python } from 'pythonia'
const np = await python('numpy')
const plot = await python('matplotlib.pyplot')

// Fixing random state for reproducibility
await np.random.seed(19680801)
const [mu, sigma] = [100, 15]
// Inline expression evaluation for operator overloading
const x = await py`${mu} + ${sigma} * ${np.random.randn(10000)}`

// the histogram of the data
const [n, bins, patches] = await plot.hist$(x, 50, { density: true, facecolor: 'g', alpha: 0.75 })
console.log('Distribution', await n) // Always await for all Python access
await plot.show()
python.exit()

通过CommonJS(没有顶级await):

const { py, python } = require('pythonia')
async function main() {
  const np = await python('numpy')
  const plot = await python('matplotlib.pyplot')
  ...
  // the rest of the code
}
main().then(() => python.exit()) // If you don't call this, the process won't quit by itself.

Boa很适合您的需求，请参阅扩展Python tensorflow keras的示例。JavaScript中的顺序类。

const fs = require('fs');
const boa = require('@pipcook/boa');
const { tuple, enumerate } = boa.builtins();

const tf = boa.import('tensorflow');
const tfds = boa.import('tensorflow_datasets');

const { keras } = tf;
const { layers } = keras;

const [
  [ train_data, test_data ],
  info
] = tfds.load('imdb_reviews/subwords8k', boa.kwargs({
  split: tuple([ tfds.Split.TRAIN, tfds.Split.TEST ]),
  with_info: true,
  as_supervised: true
}));

const encoder = info.features['text'].encoder;
const padded_shapes = tuple([
  [ null ], tuple([])
]);
const train_batches = train_data.shuffle(1000)
  .padded_batch(10, boa.kwargs({ padded_shapes }));
const test_batches = test_data.shuffle(1000)
  .padded_batch(10, boa.kwargs({ padded_shapes }));

const embedding_dim = 16;
const model = keras.Sequential([
  layers.Embedding(encoder.vocab_size, embedding_dim),
  layers.GlobalAveragePooling1D(),
  layers.Dense(16, boa.kwargs({ activation: 'relu' })),
  layers.Dense(1, boa.kwargs({ activation: 'sigmoid' }))
]);

model.summary();
model.compile(boa.kwargs({
  optimizer: 'adam',
  loss: 'binary_crossentropy',
  metrics: [ 'accuracy' ]
}));

完整的示例在:https://github.com/alibaba/pipcook/blob/master/example/boa/tf2/word-embedding.js

我在另一个项目pipook中使用了Boa，这是为了解决JavaScript开发人员的机器学习问题，我们通过Boa库在Python生态系统(tensorflow,keras,pytorch)上实现了ML/DL模型。