请教大侠们，使用fixtwo模型对面板数据进行检验时，为什么回归系数会缺失啊？以下为sas的输出结果：

 Model Description

                                 Estimation Method             FixTwo
                                 Number of Cross Sections          11
                                 Time Series Length              1454

                                            Fit Statistics

                          SSE           6021.8166    DFE               14287
                          MSE              0.4215    Root MSE         0.6492
                          R-Square         0.9011

                                     F Test for No Fixed Effects

                              Num DF      Den DF    F Value    Pr > F

                                  1463       14286      88.96    <.0001

                                         Parameter Estimates

                                        Standard
      Variable        DF    Estimate       Error    t Value    Pr > |t|    Label

      CS1              1    -0.00741      0.0241      -0.31      0.7583    Cross Sectional
                                                                           Effect    1
      CS2              1    -0.00709      0.0241      -0.29      0.7685    Cross Sectional
                                                                           Effect    2
      CS3              1    -0.02059      0.0241      -0.86      0.3924    Cross Sectional
                                                                           Effect    3

      ........

      .........

     Intercept        1    -0.87949      0.1964      -4.48      <.0001    Intercept
       cosd             0           0                0                     .          .

     上述自变量cosd的回归系数缺失。在log中显示NOTE: The transformed regression does not have full rank. User should be aware of possible  multicollinearity and/or estimability problems before using FixTwo method results.

     请问，为什么会出现这种情况啊？ 我的样本是已经按照个体单位和时间单位排列好的。而在使用fixone模型时，并不会出现这种问题。