Data Subset Selection Grouping Error

[ppines]

I am trying to create a subset - grouped by Country for 2017. Using the variables OPEC, EU, OECD, Year and Country, I’d like to highlight the countries for OPEC, EU and OECD variables. Owing to the size of the data frame I am only providing screenshots:

Here’s a sample of what the subset data frame should look like:

But my subset isn’t working:

# Group based data subset - group by Country for 2017
Global_Energy <- subset.data.frame(BP_Stats_Data,Year == 2017, select = c(Country,OPEC,EU,OECD)) 
               group_by(Country) 

Error in group_by(Country) : object 'Country' not found

Here’s a screenshot of the BP_Stats_Data:

Why can’t it find the variable in the data frame?

[andresrcs]

@ppines

group_by(Country) 
Error in group_by(Country) : object 'Country' not found

That error is happening because you are not providing a .data argument to the dplyr::group_by() function, also, using this command without a subsequent summarise() or mutate() step is going to have no effect in your dataframe.

This would be the dplyr equivalent of your subset code

library(dplyr)

Global_Energy <- BP_Stats_Data %>% 
    filter(Year == 2017) %>% 
    select(Country,OPEC,EU,OECD)
    # I'm not including the group_by() statement because I don't understand what are you trying to do with it.

Note: Even if you still can't succesfully make a subset of your dataset to put into a reprex, please include the library calls in your code, by not doing this, you make things harder for people trying to help you.

[ppines]

Thanks for the feedback. Here's the code:

Libraries

library(dplyr)
library(ggplot2)
library(tidyr)
library(ggmap)

Import the BP-Stats data into a data table called “BP_Stats_Data”

BP_Stats_Data<- read.table("bp-stats.csv", header=TRUE, sep = ",", dec= ".")

Check structure

str(BP_Stats_Data)

Display summary

summary(data.frame(BP_Stats_Data))

Remove missing data (data cleaning)

BP_Stats_Data <- na.omit(BP_Stats_Data)

Visualise levels of CO2 over the 27 years using a bar plot (data represention)

table(BP_Stats_Data$co2_mtco2,BP_Stats_Data$Year)
ggplot(BP_Stats_Data, aes(x=Year, co2_mtco2)) + geom_bar(stat="identity")+ ggtitle("Carbon dioxide levels 1990-2017")

Data subset selection of Renewables

Using subset function

Renewables <- c("biofuels_kboed", "biofuels_ktoe", "biogeo_mtoe", "biogeo_twh", "biomass_mtoe", "biomass_twh", "geothermal_mtoe", "geothermal_twh", "hydro_mtoe", "hydro_twh", "renewables_mtoe", "renewables_twh", "solar_mtoe", "solar_twh", "wind_mtoe", "wind_twh")
Renewables <- subset(BP_Stats_Data,biofuels_kboed > 0 | biofuels_ktoe > 0 | biogeo_mtoe > 0 | biogeo_twh > 0 | biomass_mtoe > 0 | biomass_twh > 0 | geothermal_mtoe > 0 | geothermal_twh > 0 | hydro_mtoe > 0 | hydro_twh > 0 | renewables_mtoe > 0 | renewables_twh > 0 | solar_mtoe > 0 | solar_twh > 0 | wind_mtoe > 0 | wind_twh > 0, select=c(Country, Year, biofuels_kboed, biofuels_ktoe, biogeo_mtoe, biogeo_twh, biomass_mtoe, biomass_twh, geothermal_mtoe, geothermal_twh, hydro_mtoe, hydro_twh, renewables_mtoe, renewables_twh, solar_mtoe, solar_twh, wind_mtoe, wind_twh))
Renewables <- Renewables %>% mutate (Combined_Renewables = rowSums(.[3:18]))
str(Renewables)
summary(Renewables)

Use ggplot2 to visualize Combined Renewables by Year

ggplot(data=Renewables, mapping = aes(x = Year,Combined_Renewables )) + geom_line(aes(col = Combined_Renewables)) + ggtitle("Combined Renewables by Year")

Data subset selection of Non-Renewables

Using subset function

Non_Renewables <- c("co2_mtco2","coalcons_mtoe","coalprod_mt","coalprod_mtoe","elect_twh","electbyfuel_coal","electbyfuel_gas","electbyfuel_oil","electbyfuel_other","gascons_bcfd","gascons_bcm","gascons_mtoe","gasprod_bcfd","gasprod_bcm","gasprod_mtoe","nuclear_mtoe","nuclear_twh","oilcons_kbd","oilcons_mt","oilcons_mtoe","oilprod_kbd","oilprod_mt","oilreserves_bbl","primary_mtoe")
Non_Renewables <- subset(BP_Stats_Data,co2_mtco2 > 0 | coalcons_mtoe > 0 | coalprod_mt > 0 | coalprod_mtoe > 0 | elect_twh > 0 | electbyfuel_coal > 0 | electbyfuel_gas > 0 | electbyfuel_oil > 0 | electbyfuel_other > 0 | gascons_bcfd > 0 | gascons_bcm > 0 | gascons_mtoe > 0 | gasprod_bcfd > 0 | gasprod_bcm > 0 | gasprod_mtoe > 0 | nuclear_mtoe > 0 | nuclear_twh > 0 | oilcons_kbd > 0 | oilcons_mt > 0 | oilcons_mtoe > 0 | oilprod_kbd > 0 | oilprod_mt > 0 | oilreserves_bbl > 0 | primary_mtoe> 0, select=c(Country, Year, co2_mtco2, coalcons_mtoe, coalprod_mt, coalprod_mtoe, elect_twh, electbyfuel_coal, electbyfuel_gas, electbyfuel_oil, electbyfuel_other, gascons_bcfd, gascons_bcm, gascons_mtoe, gasprod_bcfd, gasprod_bcm, gasprod_mtoe, nuclear_mtoe, nuclear_twh, oilcons_kbd, oilcons_mt, oilcons_mtoe, oilprod_kbd, oilprod_mt, oilreserves_bbl, primary_mtoe))
Non_Renewables <- Non_Renewables %>% mutate (Combined_Non_Renewables = rowSums(.[3:26]))
str(Non_Renewables)
summary(Non_Renewables)

Use ggplot2 to visualize Combined Non-Renewables by Year

ggplot(data=Non_Renewables, mapping = aes(x = Year,Combined_Non_Renewables )) + geom_line(aes(col = Combined_Non_Renewables)) + ggtitle("Combined Non-Renewables by Year")

Group based data subset - group by Country for 2017

Global_Energy <- subset.data.frame(BP_Stats_Data,Year == 2017, select = c(Country,OPEC,EU,OECD))
group_by(Country)
summary(Global_Energy)
str(Global_Energy)
dput(Global_Energy)

[ppines]

I amended the last piece of code to your suggestion, and it does remove the error, however, it leaves me with 10 observations:

It should give me 101 observations (all countries). I am trying to create a world map that depicts a subset for 2017. Here's what I am aiming for:

[andresrcs]

@ppines

It should give me 101 observations (all countries).

We can't know why this is happening with your specific data because we don't have access to it, could you share the csv file using some cloud storage service like Dropbox for example?

Also, please narrow down your code to the part that is relevant for your original question, this makes it easier for readers to understand your issue and improves your chances of getting help.

[ppines]

Here's the link to the source data (xlsx file):

https://www.bp.com/en/global/corporate/energy-economics/statistical-review-of-world-energy.html

Here's the cut down code as per your suggestion:

# Libraries
library(dplyr)
library(ggplot2)
library(tidyr)
library(ggmap)

Global_Energy <- BP_Stats_Data %>% 
                 filter(Year == 2017) %>% 
                 select(Country,OPEC,EU,OECD)

[andresrcs]

@ppines

Here's the link to the source data (xlsx file):

That is a link to a website not to a specific file (I don't know which one to download ), also, in your previous post, you are reading the data from a csv files not from a xlsx file.

BP_Stats_Data<- read.table("bp-stats.csv", header=TRUE, sep = ",", dec= ".")

[ppines]

@ppines

source data (xlsx file)

Ok, I don't have dropbox so I thought to send you to the source file, which I converted into a csv before uploading. In any case, you have helped me out with the initial error and I have since figured out why there were too few observations. The data cleanse had removed some of the records I need to access, so I have fixed this error. I'll close this query and thanks again for all your help.