论文的一个模型需要进行参考,但是只有scala的文件,没办法用,自己也是一个python新人,能把python代码看懂都很不错了,更不要说去研究java,请大家帮忙,帮我把这段代码转成python的,非常感谢。
class MultiSenseSkipGramEmbeddingModel(override val opts: EmbeddingOpts) extends MultiSenseWordEmbeddingModel(opts) {
val negative = opts.negative.value
val window = opts.window.value
val rng = new util.Random
val sample = opts.sample.value.toDouble
override def process(doc: String): Int = {
// given a document, below line splits by space and converts each word to Int (by vocab.getId) and filters out words not in vocab
// id of a word is its freq-rank in the corpus
var sen = doc.stripLineEnd.split(‘ ‘).map(word => vocab.getId(word.toLowerCase())).filter(id => id != -1)
val wordCount = sen.size
回答
和 http://ask.csdn.net/questions/1095307 差不多
主要是函数、lambda 的格式修改下,new util.Random,这个在 python 用 random 代替
采纳之前的回答,再详细回答这个问题