~vonfry/cpipc-2020

5286b92d10e02f4f3902279c0e912b9e8ea5f8fa — Vonfry 2 years ago 46e3745
R: add draw plot
1 files changed, 25 insertions(+), 0 deletions(-)

M 2predict.r
M 2predict.r => 2predict.r +25 -0
@@ 41,3 41,28 @@ train_and_test(df, 'pca_24')
df2 <- read.xlsx("./data/主成分数据(1).xlsx", cols = seq(3, 25), sheet = "主成分数据表1")

train_and_test(df2, 'pca_k')

# draw plot with net
# c("S-ZORB.PC_3301.DACA",
#   "S-ZORB.FC_1102.PV",
#   "S-ZORB.TC_1607.DACA",
#   "S-ZORB.AT-0004.DACA.PV",
#   "S-ZORB.FT_3702.DACA",
#   "饱和烃",
#   "S-ZORB.AT-0003.DACA.PV",
#   "S-ZORB.TXE_2203A.DACA",
#   "S-ZORB.FT_2502.DACA",
#   "S-ZORB.AT_6201.DACA",
#   "S-ZORB.PC_3101.DACA",
#   "S-ZORB.FT_1001.PV",
#   "S-ZORB.AT-0001.DACA.PV",
#   "S-ZORB.AT-0009.DACA.PV",
#   "S-ZORB.TXE_2202A.DACA",
#   "S-ZORB.PDT_3601.DACA",
#   "S-ZORB.FC_2801.PV") -> feature
# df[, feature] -> df_
#
# colnames(df_) <- paste("D", seq(1,17), sep='')
# df_[, "RONL"] <- processed["RON损失(不是变量)"]
# net_ <- neuralnet(RONL ~ ., df_, hidden = 10, rep = 100, linear.output = T)
# plot(net_)