Ian Lin

Yes, please drop Python 2.7.

You can try this: ``` stocks = [Stock(symbol) for symbol in ['2330', '6515']] ``` I don't think it's a good idea to allow `Stock(['2330', '6515'])`. Maybe you can elaborate more...

Yeah, it's very easy to be banned by TWSE. I always add a 1~5s random sleep between each request.

我有想過這個問題，不過沒有測試過 另外 user agent 可能也有影響 不知道有沒有人測試過？

`moving_average` 有問題，不能直接用 `sum()`，因為原始 data `--` 的部份會被換成 `None`。應該要先拿掉 `None` 再計算。

@Coldayftw 資料是有正常讀到的，只是分析的時候出問題。另外，圖片可以直接傳上 github。

可以參考 `pandas.DataFrame.mean`: https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.mean.html 用 `skipna` 或是 `skip_na` 等等之類的參數。 我覺得這種 data 裡面有 NA 是很合理的，不算是 error。只要能定義出一個有意義的分析就可以了。

While calling `TWSEFetcher.fetch`, I got `[Errno 101] Network is unreachable`, and my browser cannot open `http://www.tse.com.tw/`. It's a typical ban when you request TWSE too frequently.

@jpuigcerver This seems like a duplicate of #2, and [you said](https://github.com/jpuigcerver/openfst-python/issues/2#issuecomment-709994199) that: > Yes. The Python interface for OpenFst has been broken at least since version 1.7.1. Last time I...

`DataGenerator.remove(key)` shuold remove all the data that depend on `key`. If `@will_generate(handler, keys, force=True)`, then `keys` and the data that depend on `key` shuold be generated no matter it exists...