R/confin.Main.r

In halka9000stg/MyRStatLib: Library of R language on Statics, for Myself.

confin.names=function(vec){
  confin.names.proto(vec,c("lower","upper"))
}
confin.names.proto=function(vec,lab){
  names(vec)=lab
  vec
}

# (y*100)% interval estimation for a population mean of (vec), with normal-distribution.
# population variance known
confin.mean_kn=function(vec,y,pop_var){
  len=length(vec)
  mn=mean(vec)
  vr=var(vec)
  alpha=1-y
  al_hf=alpha/2
  normval_hf=qnorm(al_hf)
  lower=mn-normval_hf*sqrt(pop_var/len)
  upper=mn+normval_hf*sqrt(pop_var/len)
  res = confin.names(confin.swap(c(lower,upper)))
  res
}
confin.mean_kn.data=function(vec,y,pop_var){
  len=length(vec)
  mn=mean(vec)
  vr=var(vec)
  alpha=1-y
  al_hf=alpha/2
  normval_hf=qnorm(al_hf)
  lower=mn-normval_hf*sqrt(pop_var/len)
  upper=mn+normval_hf*sqrt(pop_var/len)
  res = confin.names.proto(c(len,mn,vr,alpha,al_hf,normval_hf),confin.mean_kn.data.returns())
  res
}
confin.mean_kn.data.returns=function(){
  c("len","mn","vr","alpha","al_hf","normval_hf")
}
# (y*100)% interval estimation for a population mean of (vec), with t-distribution.
# population variance unknown
confin.mean=function(vec,y){
  len=length(vec)
  mn=mean(vec)
  vr=var(vec)
  dof=len-1
  alpha=1-y
  al_hf=alpha/2
  tval_hf=qt(al_hf,dof)
  lower=mn-tval_hf*sqrt(vr/len)
  upper=mn+tval_hf*sqrt(vr/len)
  res = confin.names(confin.swap(c(lower,upper)))
  res
}
confin.mean.data=function(vec,y){
  len=length(vec)
  mn=mean(vec)
  vr=var(vec)
  dof=len-1
  alpha=1-y
  al_hf=alpha/2
  tval_hf=qt(al_hf,dof)
  lower=mn-tval_hf*sqrt(vr/len)
  upper=mn+tval_hf*sqrt(vr/len)
  res = confin.names.proto(c(len,mn,vr,dof,alpha,al_hf,tval_hf),confin.mean.data.returns())
  res
}
confin.mean.data.returns=function(){
  c("ln","mn","vr","dof","alpha","al_hf","tval_hf")
}
confin.mean2=function(len,mn,vr,y){
  dof=len-1
  alpha=1-y
  al_hf=alpha/2
  tval_hf=qt(al_hf,dof)
  lower=mn-tval_hf*sqrt(vr/len)
  upper=mn+tval_hf*sqrt(vr/len)
  res = confin.names(confin.swap(c(lower,upper)))
  res
}

# (y*100)% interval estimation for a population variance of (vec), with chi-distribution.
confin.var=function(vec,y){
  len=length(vec)
  mn=mean(vec)
  vr=var(vec)
  dof=len-1
  alpha=1-y
  al_hf=alpha/2
  al_hf_m=1-al_hf
  chival_hf=qchisq(al_hf,dof)
  chival_hf_m=qchisq(al_hf_m,dof)
  lower=(dof*vr)/chival_hf
  upper=(dof*vr)/chival_hf_m
  res = confin.names(confin.swap(c(lower,upper)))
  res
}
confin.var.data=function(vec,y){
  len=length(vec)
  mn=mean(vec)
  vr=var(vec)
  dof=len-1
  alpha=1-y
  al_hf=alpha/2
  al_hf_m=1-al_hf
  chival_hf=qchisq(al_hf,dof)
  chival_hf_m=qchisq(al_hf_m,dof)
  lower=(dof*vr)/chival_hf
  upper=(dof*vr)/chival_hf_m
  res = confin.names.proto(c(len,mn,dof,alpha,al_hf,al_hf_m,chival_hf,chival_hf_m),confin.var.data.returns())
  res
}
confin.var.data.returns=function(){
  c("len","mn","dof","alpha","al_hf","al_hf_m","chival_hf","chival_hf_m")
}


# (y*100)% interval estimation for a population mean difference of (vec), with normal-distribution.
# var known
confin.mean_df_kn=function(vec,y){}

# (y*100)% interval estimation for a population mean difference of (vec), with t-distribution, and mearged variance.
# var1=var2=var,but unknown
confin.mean_df_eq=function(x_vec,y_vec,y){
  x_len=10#length(x_vec)
  x_mean=102.8#mean(x_vec)
  x_var=2#var(x_vec)
  y_len=10#length(x_vec)
  y_mean=101.4#mean(x_vec)
  y_var=4#var(x_vec)
  x_dof=x_len-1
  y_dof=y_len-1
  alpha=1-y
  tval=qt(alpha,x_dof+y_dof)
  mgd_var=(x_dof*x_var+y_dof*y_var)/(x_dof+y_dof)
  lower=(x_mean-y_mean)-tval*(x_dof+y_dof)*sqrt(mgd_var*((1/x_len)+(1/y_len)))
  upper=(x_mean-y_mean)+tval*(x_dof+y_dof)*sqrt(mgd_var*((1/x_len)+(1/y_len)))
  res = confin.names(confin.swap(c(lower,upper)))
  t=((x_mean - y_mean)-(x_var - y_var))/(mgd_var*sqrt((1/x_len)+(y_len)))
  res
}
confin.mean_df_eq.data=function(x_len,x_mean,x_var,y_len,y_mean,y_var,y){
  x_dof=x_len-1
  y_dof=y_len-1
  alpha=1-y
  tval=qt(alpha,x_dof+y_dof)
  mgd_var=(x_dof*x_var+y_dof*y_var)/(x_dof+y_dof)
  lower=(x_mean-y_mean)-tval*(x_dof+y_dof)*sqrt(mgd_var*((1/x_len)+(1/y_len)))
  upper=(x_mean-y_mean)+tval*(x_dof+y_dof)*sqrt(mgd_var*((1/x_len)+(1/y_len)))
  res = confin.names.proto(c(x_dof,y_dof,alpha,tval,mgd_var),confin.mean_df_eq.data.returns())
  res
}
confin.mean_df_eq.data.returns=function(){
  c("x_dof","y_dof","alpha","tval","mgd_var")
}
# (y*100)% interval estimation for a population mean difference of (vec), with ?-distribution.
#var1!=var2,and unknown
confin.mean_df=function(vec,y){}

confin.eq_pop_var=function(x_vec,y_vec,y){
  x_len=length(x_vec)
  y_len=length(y_vec)
  x_mean=mean(x_vec)
  y_mean=mean(y_vec)
  x_var=var(x_vec)
  y_var=var(y_vec)
  x_dof=x_len-1
  y_dof=y_len-1
  alpha=1-y
  al_hf=alpha/2
  fval_hf=qf(al_hf,x_dof,y_dof)
  fval_hf_m=1/fval_hf
  lower=fval_hf_m*(x_var/y_var)
  upper=fval_hf*(x_var/y_var)
  res = confin.names(confin.swap(c(lower,upper)))
  res
}
confin.eq_pop_var.data=function(x_vec,y_vec,y){
  x_len=length(x_vec)
  y_len=length(y_vec)
  x_mean=mean(x_vec)
  y_mean=mean(y_vec)
  x_var=var(x_vec)
  y_var=var(y_vec)
  x_dof=x_len-1
  y_dof=y_len-1
  alpha=1-y
  al_hf=alpha/2
  fval_hf=qf(al_hf,x_dof,y_dof)
  fval_hf_m=1/fval_hf
  lower=fval_hf_m*(x_var/y_var)
  upper=fval_hf*(x_var/y_var)
  res = confin.names.proto(c(x_len,y_len,x_mean,y_mean,x_var,y_var,x_dof,y_dof,alpha,al_hf,fval_hf,fval_hf_m),confin.eq_pop_var.data.returns())
  res
}
confin.eq_pop_var.data.returns=function(){
  c("x_len","y_len","x_mean","y_mean","x_var","y_var","x_dof","y_dof","alpha","al_hf","fval_hf","fval_hf_m")
}


confin.ratio=function(spe_len,spe_tr,y){
  alpha=1-y
  al_hf=alpha/2
  ratio = spe_tr/spe_len
  normval=qnorm(al_hf)
  lower = ratio - normval*sqrt((ratio*(1-ratio))/spe_len)
  upper = ratio + normval*sqrt((ratio*(1-ratio))/spe_len)
  res = confin.names(confin.swap(c(lower,upper)))
  res
}
confin.nsur=function(y,max_wid){
  alpha=1-y
  al_hf=alpha/2
  normval=qnorm(al_hf)
  nsq=normval*(1/max_wid)
  nsq^2
}




confin.swap=function(vec){
  if(length(vec)!=2){
    warning("length of vecter in argument must be 2")
  }
  a=vec[1]
  b=vec[2]
  if(a>b){
    return(c(b,a))
  }else{
    return(vec)
  }
}







# confin=list("mean_kn"=function(vec,y,pop_var){confin.mean_kn(vec,y,pop_var)},
#             "mean_kn.data"=function(vec,y,pop_var){confin.mean_kn.data(vec,y,pop_var)},
#             "mean_kn.data.returns"=function(){confin.mean_kn.data.returns()},
#             "mean"=function(vec,y){confin.mean(vec,y)},
#             "mean.data"=function(vec,y){confin.mean.data(vec,y)},
#             "mean.data.returns"=function(){confin.mean.data.returns()},
#             "mean2"=function(len,mn,vr,y){confin.mean2(len,mn,vr,y)},
#             "var"=function(vec,y){confin.var(vec,y)},
#             "var.data"=function(vec,y){confin.var.data(vec,y)},
#             "var.data.returns"=function(){confin.var.data.returns()},
#             "mean_df_kn"=function(vec,y){confin.mean_df_kn(vec,y)},
#             "mean_df_eq"=function(){confin.mean_df_eq},
#             "mean_df_eq.data"=function(x_len,x_mean,x_var,y_len,y_mean,y_var,y){confin.mean_df_eq.data(x_len,x_mean,x_var,y_len,y_mean,y_var,y)},
#             "mean_df_eq.data.returns"=function(){confin.mean_df_eq.data.returns()},
#             "eq_pop_var"=function(x_vec,y_vec,y){confin.eq_pop_var(x_vec,y_vec,y)},
#             "eq_pop_var.data"=function(x_vec,y_vec,y){confin.eq_pop_var.data(x_vec,y_vec,y)},
#             "eq_pop_var.data.returns"=function(){confin.eq_pop_var.data.returns()},)