如何在不影响文件格式的情况下更改文本文件中的行数组

Question

我有一个 DFT 代码 (VASP) 的输入文件 POSCAR，如下所示：

System
1.00000000000000
**5.2916193007999999    0.0000000000000000    0.0000000000000000
0.0000000000000000    5.2916193007999999    0.0000000000000000
0.0000000000000000    0.0000000000000000   20.0097122192000008**

       Ta   N
         2     4
    Direct
      0.0000000000000000  0.0000000000000000  0.5000000000000000
      0.5000000000000000  0.5000000000000000  0.5000000000000000
      0.1430930929177486  0.6430931269177513  0.5336780085877783
      0.8569068730822487  0.3569069180822488  0.5336780085877783
      0.6430931269177513  0.8569068730822487  0.4663219914122216
      0.3569069180822488  0.1430930929177486  0.4663219914122216

我想变形，或者我想说，我想更改一些具有特定百分比的参数，比如 0.5%，（这些以粗体显示）。我们可以将粗体参数的数组命名为：

a11 a12 a13
a21 a22 a23
a31 a32 a33

根据具体情况，我需要将a11修改为a33（有时可能只有a11或a11修改为a13或全部为a）变形+0.5%，变形-0.5%

我的作业文件中需要一个 shell (bash) 脚本，以便生成所需的变形数据文件。

最终数据应该在两个目录中减去（它将有 a-a0.005）和加号（（它将有 a+a0.005）。

最终数据应该是这样的：

在plus(mkdir plus)目录下，POSCAR文件应该是这个：

System
   1.00000000000000     
     5.3180773973039998995    1.0050000000000000000    1.4925000000000000000
     1.0050000000000000000    5.3180773973039998995    0.2010000000000000000
     1.4925000000000000000    0.2010000000000000000   10.0500000000000000000
   Ta   N
     2     4
Direct
  0.0000000000000000  0.0000000000000000  0.5000000000000000
  0.5000000000000000  0.5000000000000000  0.5000000000000000
  0.1430930929177486  0.6430931269177513  0.5336780085877783
  0.8569068730822487  0.3569069180822488  0.5336780085877783
  0.6430931269177513  0.8569068730822487  0.4663219914122216
  0.3569069180822488  0.1430930929177486  0.4663219914122216

在minus（mkdir minus）目录下，POSCAR文件应该是这个：

System
   1.00000000000000     
     5.2651612042959999005    0.9950000000000000000    1.4925000000000000000
     0.9950000000000000000    5.2651612042959999005    0.1990000000000000000
     1.4925000000000000000    0.1990000000000000000   9.9500000000000000000
   Ta   N 
     2     4
Direct
  0.0000000000000000  0.0000000000000000  0.5000000000000000
  0.5000000000000000  0.5000000000000000  0.5000000000000000
  0.1430930929177486  0.6430931269177513  0.5336780085877783
  0.8569068730822487  0.3569069180822488  0.5336780085877783
  0.6430931269177513  0.8569068730822487  0.4663219914122216
  0.3569069180822488  0.1430930929177486  0.4663219914122216

重要的是，直接下面出现的所有行都不会发生变化。

我正在尝试如下使用

#!/bin/bash
DEFORM=0.005
A11=$(cat POSCAR | head -n 3 | tail -n 1 | awk '{print }')
A12=$(cat POSCAR | head -n 3 | tail -n 1 | awk '{print }')
A13=$(cat POSCAR | head -n 3 | tail -n 1 | awk '{print }')
A21=$(cat POSCAR | head -n 4 | tail -n 1 | awk '{print }')
A22=$(cat POSCAR | head -n 4 | tail -n 1 | awk '{print }')
A23=$(cat POSCAR | head -n 4 | tail -n 1 | awk '{print }')
A31=$(cat POSCAR | head -n 5 | tail -n 1 | awk '{print }')
A32=$(cat POSCAR | head -n 5 | tail -n 1 | awk '{print }')
A33=$(cat POSCAR | head -n 5 | tail -n 1 | awk '{print }')
PA11=$(echo "$A11" "$DEFORM" | awk '{printf "%.16f\n",  +*}')
sed -i "s/$A11/$PA11/g" POSCAR

但这似乎是一个艰难的过程。我正在寻找一个简单的脚本。

Answer 1

我不熟悉这种文件格式，所以你得自己检查一下。

就awk而言，创建一个脚本(poscar.awk):

BEGIN{ DEFORM=0.005 }
NR>=3 && NR<=5{ 
  =-DEFORM 
  =-DEFORM 
  =-DEFORM 
  printf "         %.16f %.16f %.16f\n", , ,    # This prints lines 3-5
}
!(NR>=3 && NR<=5)                                     # This prints all other lines

然后您可以：awk -f poscar.awk inputfile 这将导致：

System
       1.00000000000000
         5.2866193008000000 0.9950000000000000 1.4950000000000001
         3.9950000000000001 5.2866193008000000 0.1950000000000000
         2.9950000000000001 0.2950000000000000 9.9949999999999992
       Ta   N
         2     4
    Direct
      0.0000000000000000  0.0000000000000000  0.5000000000000000
      0.5000000000000000  0.5000000000000000  0.5000000000000000
      0.1430930929177486  0.6430931269177513  0.5336780085877783
      0.8569068730822487  0.3569069180822488  0.5336780085877783
      0.6430931269177513  0.8569068730822487  0.4663219914122216
      0.3569069180822488  0.1430930929177486  0.4663219914122216

Answer 2

这可能是您正在尝试做的事情：

$ cat tst.awk
NR==1 {
    plus  = FILENAME "_plus"
    minus = FILENAME "_minus"
}
(3 <= NR) && (NR <= 5) {
    for (i=1; i<=NF; i++) {
        printf "%0.16f%s", $i * 1.005, (i<NF ? OFS : ORS) > plus
        printf "%0.16f%s", $i * 0.095, (i<NF ? OFS : ORS) > minus
    }
    next
}
{
    print > plus
    print > minus
}

$ awk -f tst.awk file

$ head -100 file_*
==> file_minus <==
System
1.00000000000000
0.5027038335760000 0.0000000000000000 0.0000000000000000
0.0000000000000000 0.5027038335760000 0.0000000000000000
0.0000000000000000 0.0000000000000000 1.9009226608240002

       Ta   N
         2     4
    Direct
      0.0000000000000000  0.0000000000000000  0.5000000000000000
      0.5000000000000000  0.5000000000000000  0.5000000000000000
      0.1430930929177486  0.6430931269177513  0.5336780085877783
      0.8569068730822487  0.3569069180822488  0.5336780085877783
      0.6430931269177513  0.8569068730822487  0.4663219914122216
      0.3569069180822488  0.1430930929177486  0.4663219914122216

==> file_plus <==
System
1.00000000000000
5.3180773973039992 0.0000000000000000 0.0000000000000000
0.0000000000000000 5.3180773973039992 0.0000000000000000
0.0000000000000000 0.0000000000000000 20.1097607802959999

       Ta   N
         2     4
    Direct
      0.0000000000000000  0.0000000000000000  0.5000000000000000
      0.5000000000000000  0.5000000000000000  0.5000000000000000
      0.1430930929177486  0.6430931269177513  0.5336780085877783
      0.8569068730822487  0.3569069180822488  0.5336780085877783
      0.6430931269177513  0.8569068730822487  0.4663219914122216
      0.3569069180822488  0.1430930929177486  0.4663219914122216

如何在不影响文件格式的情况下更改文本文件中的行数组

How to change a array of rows in a text file without disturbing the format of the file

awk

sed