我一直在努力使用 bash 将 .csv 文件转换为所需的表格式。
我的文件o/p:
$ cat finalcount.csv
state,status,count
NW,CONGESTED,11
NW,CLEARED,9
ACT,CLEARED,7
ACT,CONGESTED,7
QLD,CLEARED,5
WA,CONGESTED,3
WA,CLEARED,2
VIC,CLEARED,3
VIC,CONGESTED,2
我想要的o/p:
State Congested Cleared
ACT 7 7
NW 11 9
VIC 2 3
QLD - 5
请分享您的答案并进行解释。
答案1
使用磨坊主:
$ mlr --icsvlite --opprint \
reshape -s status,count then unsparsify then reorder -e -f CLEARED finalcount.csv
state CONGESTED CLEARED
NW 11 9
ACT 7 7
QLD - 5
WA 3 2
VIC 2 3
GNU 数据整合对于数据透视表(又名交叉表)之类的东西也很方便,尽管似乎没有任何简单的方法来控制列顺序:
datamash --header-in --filler=- -t, crosstab 1,2 unique 3 < finalcount.csv | csvformat -T
CLEARED CONGESTED
ACT 7 7
NW 9 11
QLD 5 -
VIC 3 2
WA 2 3
如果你确实需要使用 Awk,那么这样的东西应该可以工作:
awk -F, '
NR>1 {s[$1]}
$2 == "CONGESTED" {a[$1]=$3}
$2 == "CLEARED" {b[$1]=$3}
END {
printf "state\tcongested\tcleared\n";
for (k in s)
printf "%s\t%s\t%s\n", k, a[k]=="" ? "-" : a[k], b[k]=="" ? "-" : b[k]
}
' finalcount.csv
答案2
$ cat tst.awk
BEGIN { FS=","; OFS="\t" }
{ sub(/\r$/,"") }
NR > 1 {
states[$1]
statuses[$2]
vals[$1,$2] = $NF
}
END {
printf "%s", "state"
for (status in statuses) {
printf "%s%s", OFS, status
}
print ""
for (state in states) {
printf "%s", state
for (status in statuses) {
printf "%s%s", OFS, ((state,status) in vals ? vals[state,status] : "-")
}
print ""
}
}
。
$ awk -f tst.awk file
state CONGESTED CLEARED
QLD - 5
ACT 7 7
WA 3 2
NW 11 9
VIC 2 3
。
$ awk -f tst.awk file | column -s $'\t' -t
state CONGESTED CLEARED
QLD - 5
ACT 7 7
WA 3 2
NW 11 9
VIC 2 3