There are two syntaxes for setting the data:

                tsset timevar

                tsset panelvar timevar

    In the first syntax -- tsset timevar -- the data are set to be a straight time series.

    In the second syntax -- tsset panelvar timevar -- the data are set to be a collection of time series, one for
    each value of panelvar, also known as panel data, cross-sectional time-series data, and xt data.  Such datasets
    can be analyzed by xt commands as well as ts commands.  If you tsset panelvar timevar, you do not need to xtset
    panelvar timevar to use the xt commands.

如果是要处理panel data，两者等价；
如果处理费panel data，参见详细命令

处理非面板数据的话，能不能介绍一下区别，我看不太懂

处理非面板数据
xtset is used to declare the dataset is panel data, see example

webuse pig
xtset id week

declare data is panel with id as identifier and stub is week
.        xtset id week
        panel variable:        id (strongly        balanced)
        time variable:        week, 1 to 9
        delta:        1 unit

tsset is used to declare the data is time-series data, see example
use the same dataset as above:

. tsset week
repeated time        values        in        sample
r(451);

the reason of error is because week is not a time-series variable. list  week in 1/10

     | week |
     |------|
  1. |    1 |
  2. |    1 |
  3. |    1 |
  4. |    1 |
  5. |    1 |
     |------|
  6. |    1 |
  7. |    1 |
  8. |    1 |
  9. |    1 |
10. |    1 |


now show a case where tsset can be applied to time-series data
webuse dow1
tsset date



.             . tsset date
        time variable:  date, 02jan1953 to 20feb1990, but with gaps
                delta:  1 day

list date in 1/10

      |      date |
      |-----------|
   1. | 02jan1953 |
   2. | 05jan1953 |
   3. | 06jan1953 |
   4. | 07jan1953 |
   5. | 08jan1953 |
      |-----------|
   6. | 09jan1953 |
   7. | 12jan1953 |
   8. | 13jan1953 |
   9. | 14jan1953 |
  10. | 15jan1953 |

learning.