Forecasting Interval Valued Stock Returns Based on News Media Sentiments_data and code

Data and code are used to analyze the data and results in the article “Forecasting Interval Valued Stock Returns Based on News Media Sentiments.”1. The sample period of the basic data is the weekly data from 2014/01/30 to 2022/10/14.2, data.csv and data_index.xlsx are the basic data sets, for the subsequent results of the foundation of the analysis; the notation sz, yysw, fyjr, yh, dz, jsj, nlmy in table represent the SZSE index, pharmaceuticals and biologicals, non-banking financial, banking, electronics, computers, and agriculture, forestry, animal husbandry and fisheries industry, respectively, corresponding to the content of article. The rest of the data files are derived from the analysis process and used for analysis, corresponding to the graphics and tables in our article, see the comments in our code file for details.3、In our table, the notation "max" and "min" represent the sequence of maximum and minimum values; "p" represents the sequence of price index, "D" represents the sequence of difference, and "ln" represents the sequence of logarithm.4. The code is in R language, corresponding to the results of the charts in the text, see the comments in the code.