Awk If语句跳过零除

我正试图在一个类似这样的文件中跳过零除法。对于col1中的每个元素，求col3和col4中的值之和，然后除以sumcol3/sumcol4

A 3 0.100000 0.200000
A 4 0.100000 0.200000
B 1 0.200000 0.200000
B 5 0.200000 0.000000
C 3 0.200000 0.200000
X 3 0.000000 0.000000
X 5 0.200000 0.200000
D 100 0.00000 0.200000

# set fields delimiters 
BEGIN { FS = OFS = "\t" }

# the first time col1 value occurs, store other cols
!h[$1] {
    h[$1] = ++n  # save ordering
    d[n] = $1 OFS $2
}

# store sums
{
    i = h[$1]  # recover ordering
    a[i] += $3
    b[i] += $4
}

# output cols and the computed values
END {
    for (i=1; i<=n; i++) {if (b[i]!="0.000000") {print d[i], a[i], b[i], a[i]/b[i]}}
}

调整以下代码

# set fields delimiters 
BEGIN { FS = OFS = "\t" }

# the first time col1 value occurs, store other cols
!h[$1] {
    h[$1] = ++n  # save ordering
    d[n] = $1 OFS $2
}

# store sums
{
    i = h[$1]  # recover ordering
    a[i] += $3
    b[i] += $4
}

# output cols and the computed values
END {
    for (i=1; i<=n; i++) {if (b[i]!="0.000000") {print d[i], a[i], b[i], a[i]/b[i]}}
}

---

在awk中，在算术上下文中计算为零的表达式总是被视为

0

1。也就是说，当

b[i]

的值为零时

# set fields delimiters 
BEGIN { FS = OFS = "\t" }

# the first time col1 value occurs, store other cols
!h[$1] {
    h[$1] = ++n  # save ordering
    d[n] = $1 OFS $2
}

# store sums
{
    i = h[$1]  # recover ordering
    a[i] += $3
    b[i] += $4
}

# output cols and the computed values
END {
    for (i=1; i<=n; i++) {if (b[i]!="0.000000") {print d[i], a[i], b[i], a[i]/b[i]}}
}

b[i] != "0.000000"

同：

# set fields delimiters 
BEGIN { FS = OFS = "\t" }

# the first time col1 value occurs, store other cols
!h[$1] {
    h[$1] = ++n  # save ordering
    d[n] = $1 OFS $2
}

# store sums
{
    i = h[$1]  # recover ordering
    a[i] += $3
    b[i] += $4
}

# output cols and the computed values
END {
    for (i=1; i<=n; i++) {if (b[i]!="0.000000") {print d[i], a[i], b[i], a[i]/b[i]}}
}

0 != "0.000000"

如果计算结果为

true

，则不明确。因此，您应该使用

b[i]=0或删除
=“0.000000”
2

# set fields delimiters 
BEGIN { FS = OFS = "\t" }

# the first time col1 value occurs, store other cols
!h[$1] {
    h[$1] = ++n  # save ordering
    d[n] = $1 OFS $2
}

# store sums
{
    i = h[$1]  # recover ordering
    a[i] += $3
    b[i] += $4
}

# output cols and the computed values
END {
    for (i=1; i<=n; i++) {if (b[i]!="0.000000") {print d[i], a[i], b[i], a[i]/b[i]}}
}


1尝试运行以下命令：

# set fields delimiters 
BEGIN { FS = OFS = "\t" }

# the first time col1 value occurs, store other cols
!h[$1] {
    h[$1] = ++n  # save ordering
    d[n] = $1 OFS $2
}

# store sums
{
    i = h[$1]  # recover ordering
    a[i] += $3
    b[i] += $4
}

# output cols and the computed values
END {
    for (i=1; i<=n; i++) {if (b[i]!="0.000000") {print d[i], a[i], b[i], a[i]/b[i]}}
}

awk 'BEGIN{print 0.0 + 0.000000 + 0}'

2在awk和许多其他脚本/编程语言中，计算为非零值的表达式在布尔上下文中计算为
true
。所以，
b[i]=当
b[i]
持有一个数值时，0和
b[i]
是相同的。
我没有测试它，因为您的样本不是正确的输入文件（例如-->$19不在那里）。我想我会采取这种方法。我已经添加了一个警告语句来获取输出中的警告，如果您不需要它，可以将其删除

# set fields delimiters 
BEGIN { FS = OFS = "\t" }

# the first time col1 value occurs, store other cols
!h[$1] {
    h[$1] = ++n  # save ordering
    d[n] = $1 OFS $2
}

# store sums
{
    i = h[$1]  # recover ordering
    a[i] += $3
    b[i] += $4
}

# output cols and the computed values
END {
    for (i=1; i<=n; i++) {if (b[i]!="0.000000") {print d[i], a[i], b[i], a[i]/b[i]}}
}

您可以将
的

循环行更改为以下内容

# set fields delimiters 
BEGIN { FS = OFS = "\t" }

# the first time col1 value occurs, store other cols
!h[$1] {
    h[$1] = ++n  # save ordering
    d[n] = $1 OFS $2
}

# store sums
{
    i = h[$1]  # recover ordering
    a[i] += $3
    b[i] += $4
}

# output cols and the computed values
END {
    for (i=1; i<=n; i++) {if (b[i]!="0.000000") {print d[i], a[i], b[i], a[i]/b[i]}}
}

for (i=1; i<=n; i++) {printf("%d %d %d %d\n",d[i], a[i], b[i], b[i]!=0?a[i]/b[i]:"Warn: There was an attempt to divide by zero")}

---

for（i=1；iIt似乎有效。我不明白。为什么只说如果（b[i]）有效呢？Thnxs！你正在用字符串比较一个数字。这永远是真的。为了代码的清晰，你应该写
b[i]！=0
。如果（b[i]）

，你可以写

，但这会让你的代码更难读（一周后你就会忘记它的意思）。