Question

我有一个包含25个变量的数据框df。

我想使用预测包为每个变量预测一天的价值。

df # original df

df2 <- df2 <- apply(df, 2, function (x) forecast(tbats(ts(x,frequency=365)), h=1))
$GOOG.Volume
         Point Forecast    Lo 80   Hi 80  Lo 95   Hi 95
3.142466        1654547 920711.4 2765229 650538 3544211

$GOOG.Adjusted
         Point Forecast   Lo 80    Hi 80    Lo 95    Hi 95
3.142466       530.7542 521.654 540.0065 516.8974 544.9667

$Avg_vol_10
         Point Forecast   Lo 80   Hi 80   Lo 95   Hi 95
3.142466        1889756 1761678 2027142 1697429 2103866

.....

> df2$GOOG.Volume
         Point Forecast    Lo 80   Hi 80  Lo 95   Hi 95
3.142466        1654547 920711.4 2765229 650538 3544211
> df2$GOOG.Volume$mean
Time Series:
Start = c(3, 53) 
End = c(3, 53) 
Frequency = 365 
[1] 1654547
attr(,"biasadj")
[1] FALSE

如何将所有这1天的预测平均值添加到数据框df中。

谢谢

这是我当前的数据框，

sample df 我想在此数据框的末尾添加每列的第一个预测的平均值。

链接到数据框。 https://drive.google.com/open?id=1MkAHU-VcVpjQelsMOLaNoHPFwH2mbzig

最终df

col_a    col_a    col_a    col_a    col_a
all present values, 

last row, we appended  (rbind ) with forecasted means

predictions of all columns


pred_cola   pred_colb   pred_colc   pred_cold


This is my requirement.

Answer 1

这是您要做什么的一个最小示例：

# sample data, replace this with yours
df <- data.table(google=runif(n = 10, 10,100),
                 apple=runif(n=10, 10, 200),
                 shell=runif(n=10, 20, 400))

# apply forecast function
df2 <- apply(df, 2, function (x) forecast(tbats(ts(x,frequency=365)), h=1))

# get mean values in a separate vector
stock_cols <- colnames(df)
get_mean_values <- sapply(stock_cols, function(x) df2[[x]]$mean, USE.NAMES = F)

# convert mean values vector to data frame
get_mean_df <- t(data.frame(get_mean_values))
colnames(get_mean_df) <- stock_cols
row.names(get_mean_df) <- NULL

## rbind values in the original dataframe
df <- rbind(df, get_mean_df)

print(df)

     google     apple     shell
 1: 99.52935 170.18530 387.77822
 2: 46.12997  41.67531 265.41992
 3: 90.60817 155.16369  94.59444
 4: 88.91726 108.59824 155.72855
 5: 36.68363  61.81021 214.94687
 6: 21.93327  75.20860 355.53876
 7: 33.93198 143.96431  78.01367
 8: 10.51548 122.39171 107.57541
 9: 83.76416  99.18202  81.93566
10: 25.71955  73.91156 253.45181
11: 56.22397 106.50459 206.37342 ## this rows shows the mean value.

将预测值添加到原始df

1 个答案: