Why is my sum() summing to zero, having no NAs and only numerical data?-CodePudding

Once again with a very simple question.

I am trying to add all emissions together, basically summing 5 variables per row.

However, it keeps summing to zero, even when I have no NAs and only numeric values.

This is the data I am working with:

df_structure <-
  structure(
    list(
      `Particeles_PM10_[kg]_WTW_whole transport chain` = c(
        0.000440486,
        0.010753239,
        0.0005393157,
        0.0107265319,
        0.200272577,
        0.169998242
      ),
      `SO2_[kg]_WTW_whole transport chain` = c(
        0.0034873728,
        0.1861534833,
        0.01613152798,
        0.185923214,
        3.715316736,
        3.155906431
      ),
      `NOX_[kg]_WTW_whole transport chain` = c(
        0.024214311,
        0.618727269,
        0.053631226,
        0.617528662,
        12.271221,
        10.3988076
      ),
      `NMHC_[kg]_WTW_whole transport chain` = c(
        0.0043159575,
        0.0385331658,
        0.0033238124,
        0.038634107,
        0.7067915367,
        0.59608807
      )
    ),
    row.names = c(NA,-6L),
    class = c("tbl_df", "tbl", "data.frame")
  )

And heres my code:

df_structure %>%
  rowwise() %>% 
  mutate(sum_emissions = sum(as.numeric("Particeles_PM10_[kg]_WTW_whole transport chain",
                         "SO2_[kg]_WTW_whole transport chain",
                         "NOX_[kg]_WTW_whole transport chain",
                         "NMHC_[kg]_WTW_whole transport chain"), na.rm = TRUE)) 
summary(df_structure$sum_emissions)

What am I doing wrong? I can open my data.frame and every column has 5 rows of filled-in data, yet the sum keeps being 0...

Thanks in advance!

CodePudding user response：

You could use across() to select what columns you want to sum up and pass the whole across() into rowSums().

library(dplyr)

df_structure %>%
  mutate(sum_emissions = rowSums(across(everything())))

# # A tibble: 6 × 5
#   `Particeles_PM10_[kg]…` `SO2_[kg]_WTW_…` `NOX_[kg]_WTW_…` `NMHC_[kg]_WTW…` sum_emissions
#                     <dbl>            <dbl>            <dbl>            <dbl>         <dbl>
# 1                0.000440          0.00349           0.0242          0.00432        0.0325
# 2                0.0108            0.186             0.619           0.0385         0.854 
# 3                0.000539          0.0161            0.0536          0.00332        0.0736
# 4                0.0107            0.186             0.618           0.0386         0.853 
# 5                0.200             3.72             12.3             0.707         16.9   
# 6                0.170             3.16             10.4             0.596         14.3

CodePudding user response：

You need to specify that it is a vector of variables using c() and ``'s. As you output is already numeric, you won't need to specify that.

df_structure %>%
  rowwise() %>% 
  mutate(sum_emissions = sum(c(`Particeles_PM10_[kg]_WTW_whole transport chain`,
                                 `SO2_[kg]_WTW_whole transport chain`,
                                 `NOX_[kg]_WTW_whole transport chain`,
                                 `NMHC_[kg]_WTW_whole transport chain`), na.rm = TRUE)) %>%
  ungroup()

A simpler way might be to use c_across:

df_structure %>%
  rowwise() %>% 
  mutate(sum_emissions = sum(c_across(1:4), na.rm = TRUE)) %>%
  ungroup()

A base solution is to use rowSums directly (or through mutate):

df_structure$sum_emissions <- rowSums(df_structure)