Select Language

AI社区

公开数据集

bogbt 推特数据

bogbt 推特数据

0.05M
331 浏览
0 喜欢
0 次下载
0 条讨论
Earth and Nature,News,Investing,Economics,Social Networks,Email and Messaging Classification

数据结构 ? 0.05M

    Data Structure ?

    * 以上分析是由系统提取分析形成的结果,具体实际数据为准。

    README.md

    Context Data for tutorial on paper reproducibility from Ratchel T. https://www.youtube.com/watch?v=352wMM6b2oo Content The file calm_sentiment_data.csv contains Twitter sentiment data for the 'calm' sentiment, aligned with returns of the DJI (the Dow Jones Index) and GSPC (the S&P 500 index). The fields consist of: * date : the date of the observed Twitter sentiment data. The data are observable by 23:59:59 EST of the given date. * tone : the tone sentiment series. * calm : the calm sentiment series. * tone_Z10 : the normalized tone variable, using k=10. This is essentially a centered Z-score of tone over 21 observations. * tone_Z1 : the normalized tone variable, using k=1. This is essentially a centered Z-score of tone over 3 (!) observations. * DJI : the closing value of the DJI index for the given day. Will take value of NA when the market is closed. Early closes are treated as normal market days. The DJI data are source from Yahoo finance via the quantmod package. * DJI_volume_k : the volume of the DJI index for the given day, in thousands. This may be useful for data QA, for example. The volume is NA when the market is closed. * DJI_forward_return : this is a one market period relative 'return' of the DJI index. A value of 0.01, for example, corresponds to a 1% increase in the DJI index. This is a _forward_ return, meaning it is the return from the close of the given day to the close of the next market day, and could be approximately captured by a long holder in the index. (Were that possible; an index is not an ETF.) The return is NA when the market is closed. Note that there is an overlap in information between the Twitter sentiment series and the forward return: the sentiment data is only observable just before midnight EST, while one would have to invest just before 4PM EST to capture the forward return. * GSPC : the closing value of the GSPC index for the given day. Will take value of NA when the market is closed. Early closes are treated as normal market days. The GSPC data are source from Yahoo finance via the quantmod package. While the original paper does not make claims regarding Twitter mood and the GSPC series, it represents broad market returns and makes a suitable target for a putative forecast of the market by sentiment. * GSPC_volume_k : the volume of the GSPC index for the given day, in thousands. * GSPC_forward_return : the one market period relative 'return' of the GSPC index. This is a forward_return, meaning it is the return from the close of the given day to the close of the next market day. The return is NA when the market is closed. Acknowledgements https://github.com/shabbychef/bogbt
    ×

    帕依提提提温馨提示

    该数据集正在整理中,为您准备了其他渠道,请您使用

    注:部分数据正在处理中,未能直接提供下载,还请大家理解和支持。
    暂无相关内容。
    暂无相关内容。
    • 分享你的想法
    去分享你的想法~~

    全部内容

      欢迎交流分享
      开始分享您的观点和意见,和大家一起交流分享.
    所需积分:0 去赚积分?
    • 331浏览
    • 0下载
    • 0点赞
    • 收藏
    • 分享