由于eviews比较支持英文，所以我们再给变量命名的时候要用英文，但是这样就不太利于我们最变量的理解了，尤其是变量比较多的时候，很容易把变量的意义弄混，这时我们就可以借助文本对象，把每个变量的中文含义写进文本对象中，这样我们在不清楚那个变量是什么含义时就可以查看文本对象中的内容。文本对象就和我们平时的用法一样，可以在里面写文字。
我们可以通过建立一个组对象来导入数据。这样我们可以设置变量的排放顺序，而不是使用eviews中默认的按照字母顺序排序。
在已知随机变量的分布形式时，如何估计其密度函数中的参数？
比如说有一个序列，已知此序列服从极值分布，那么如何求极值分布中的参数。步骤如下： 打开序列x,然后discriptive statistics&tests---empirical distribution tests
对于参数的初始值你可以自己确定，如果你不知道，也可以让eviews帮你计算，这样你保持参数框空白就行。 点击ok键，结果如下：
从估计结果中，我们看出eviews对两个参数给出了估计结果：m=-0.866844，s=2.228739
Empirical Distribution Tests
EViews provides built-in Kolmogorov-Smirnov, Lilliefors, Cramer-von Mises, Anderson-Darling, and Watson empirical distribution
Eviews提供了built-in Kolmogorov-Smirnov, Lilliefors, Cramer-von Mises, Anderson-Darling, and Watson 五种经验分布检验。这些检验建立在经验分布和理论分布函数比较的基础上。如果你想更深入的了解经验分布的检验，可以看D'Agostino and Stephens (1986)。你可以检验某个分布是不是正态分布的，或者是不是来自其他分布比如说指数分布，极值分布，logistic分布，卡方分布，weibull分布，伽马（gamma）分布。你可以提供这些分布的参数，如果你不提供，也可以让eviews为你估计这些参数。
tests. These tests are based on the comparison between the empirical distribution and the specified theoretical distribution function. For a general description of empirical distribution function testing, see D'Agostino and Stephens (1986).
You can test whether your series is normally distributed, or whether it comes from, among others, an exponential, extreme value, logistic, chi-square, Weibull, or gamma distribution. You may provide parameters for the distribution, or EViews will estimate the parameters for you.
To carry out the test, simply double click on the series and select View/Descriptive Statistics & Tests/Empirical Distribution Tests... from the series window.
There are two tabs in the dialog. The Test Specification tab allows you to specify the parametric distribution against which you want to test the empirical distribution of the series. Simply select the distribution of interest from the drop-down menu. The small display window will change to show you the parameterization of the specified distribution.
You can specify the values of any known parameters in the edit field or fields. If you leave any field blank, EViews will estimate the corresponding parameter using the data contained in the series.
估计选项选项卡提供所需要的迭代估计的相应设置。一般情况下你不需要改变这里的默认设置，除非显示估计失败。这个选项卡里的大部分选项都应该是不言而喻的。如果你选择User-specified starting values，eviews将会把系数向量c里的值作为初始值。 The Estimation Options tab provides control over any iterative estimation that is required. You should not need to use this tab unless the output indicates failure in the estimation process. Most of the options in this tab should be self-explanatory. If you select User-specified starting values, EViews will take the starting values from the C coefficient vector.
It is worth noting that some distributions have positive probability on a restricted domain. If the series data take values outside this domain, EViews will report an out-of-range error. Similarly, some of the distributions have restrictions on domain of the parameter values. If you specify a parameter value that does not satisfy this restriction, EViews will report an error message.值得注意的是，一些分布，对于一个限制域有正的概率。 如果序列中有某些取值没有包括在这个域中，eviews就会报告out-of-range error。类似的，一些分布对参数值的域有限制。
如果你所设定的参数值不满足这个限制，eviews就会给出一个错误信息。 输出结果包括两部分，第一部分是检验统计量和相应的概率。
The output from this view consists of two parts. The first part displays the test statistics and associated probability values.
这里是一个检验序列是否服从正态分布的检验结果，其中均值和方差都是根据序列样本数据计算的。
Here, we show the output from a test for normality where both the mean and the variance are estimated from the series data. The first column, \sample correction or adjusted for parameter uncertainty (in case the parameters are estimated). The third column reports p-value for the adjusted statistics.
All of the reported EViews p-values will account for the fact that parameters in the distribution have been estimated. In cases where estimation of parameters is involved, the distributions of the goodness-of-fit statistics are non-standard and distribution dependent, so that EViews may report a subset of tests and/or only a range of p-value. In this case, for example, EViews reports the Lilliefors test statistic instead of the Kolmogorov statistic since the parameters of the normal have been estimated. Details on the computation of the test statistics and the associated p-values may be found in Anderson and Darling (1952, 1954), Lewis (1961), Durbin (1970), Dallal and Wilkinson (1986), Davis and Stephens (1989), Cs?rg? and Faraway (1996) and Stephens (1986).
The second part of the output table displays the parameter values used to compute the theoretical distribution function. Any parameters that are specified to estimate are estimated by maximum likelihood (for the normal distribution, the estimate of the standard deviation is degree of freedom corrected if the mean is not specified a priori). For parameters that do not have a closed form analytic solution, the likelihood function is maximized using analytic first and second derivatives. These estimated parameters are reported with a standard error and p-value based on the asymptotic normal distribution.