- 积分
- 452
- 贡献
-
- 精华
- 在线时间
- 小时
- 注册时间
- 2017-2-21
- 最后登录
- 1970-1-1
|
楼主 |
发表于 2017-10-28 10:18:52
|
显示全部楼层
哦哦,是这样的,我的代码是做回归和检验,回归函数用的是regCoef,只做出回归图没有提示错误,加了检验之后才出现的错误提示,现贴出检验部分程序tval = onedtond(rc@tval , dimsizes(rc)) ;t-statistic
df = onedtond(rc@nptxy, dimsizes(rc)) - 2 ;自由度
b = tval ; b must be same size as tval (and df)
b = 0.5
prob = betainc(df/(df+tval^2),df/2.0,b)
;print(prob)
prob!0 = "lat" ; name dimensions
prob!1 = "lon"
prob&lat = SST&lat ; assign coordinate values to named dimensions
prob&lon = SST&lon
res2 = True
res2@gsnDraw = False
res2@gsnFrame = False
res2@cnFillOn = True
res2@cnLinesOn = False
res2@cnLineLabelsOn = False
res2@cnInfoLabelOn = False
res2@lbLabelBarOn = False
res2@cnMonoFillPattern = False
res2@cnLevelSelectionMode = "ExplicitLevels"
res2@cnLevels = (/0.05/) ;; set to significance level
res2@cnFillPatterns = (/17,-1/)
res2@cnFillColors = (/1,0/)
res2@cnFillScaleF = 1 ; add extra density
res2@gsnAddCyclic = True
res2@gsnLeftString = ""
res2@gsnLeftString = ""
res2@gsnRightString = ""
plots_11= gsn_csm_contour(wks,prob,res2) |
|