需要帮助进行股票数据集分析

Question

I created a data frame of stock information such as the "open","high", "close", etc. I now need to calculate the performance for each bar of the stock ( each row in the dataFrame). I would like to make a new column in the dataFrame that is equal to the "Close" column of the next row - the "Close" column value of the previous row.

• performance for individual bar is Close of the next bar minus Close of the current bar

I tried splitting up the close columns values by every 2nd row and making this new close columns values into its own column. Then make a new column subtracting this second column with the first one, however they was an issue dealing with the NaN values.

df['performance'] = df.Close[2] - df.Close[1]

This made the performance for each of the 52767 rows equal to "2.5".

I would like to make a column 'performance' that does it iteratively. For example if row 0's close value is 5 and row 1's close value is 7, then row 0's performance value should be 2, and this is done for 52767 rows.

Answer 1

pandas.Series.diff()

You can use .diff() with a period of -1 to calculate a difference from the subsequent row (as opposed to the normal behavior of difference from the previous row). For example:

# Example data
df = pd.read_csv("https://vincentarelbundock.github.io/Rdatasets/csv/fpp2/goog200.csv", index_col=0).head(10)

# Calculate difference
df['performance'] = df['value'].diff(-1)

yields

    time       value  performance
1      1  392.830017     0.317932
2      2  392.512085    -4.793823
3      3  397.305908    -0.705414
4      4  398.011322    -2.478882
5      5  400.490204    -7.605530
6      6  408.095734    -8.494751
7      7  416.590485     3.586670
8      8  413.003815    -0.606048
9      9  413.609863     0.536499
10    10  413.073364          NaN