fixed row 1 header issue in diff.awk; other small adjustments in hamming and lin_reg

1 files changed, 26 insertions, 20 deletions

diff --git a/diff.awk b/diff.awk
index 6dc427d..fb93b57 100644
--- a/diff.awk
+++ b/diff.awk
@@ -4,6 +4,7 @@
 # numerical diff along columns
 
 BEGIN {
+    OFMT = "%.9g"
     sign = "[+-]?"
     decimal = "[0-9]+[.]?[0-9]*"
     fraction = "[.][0-9]*"
@@ -12,44 +13,49 @@ BEGIN {
 }
 
 NR == 1 {
-    header_nf = NF
+    ### orig columns
     for (n=1; n<=NF; n++) {
-        printf("%s", $n)
-        ($n !~ number) ?  dheader[n] = "d" $n : dheader[n] = "dcol_" n
+        ($n !~ number) ? header[n] = $n : header[n] = "col" n
+        printf(header[n])
+        ($n !~ number) ? dheader[n] = "d" $n : dheader[n] = "dcol" n
         if (n <= NF)
-            printf("%s", OFS)
+            printf(OFS)
     }
 
+    ### diff columns
     for (n=1; n<=NF; n++) {
-        printf("%s", dheader[n])
+        printf(dheader[n])
         if (n < NF)
-            printf("%s", OFS)
+            printf(OFS)
     }
 }
 
-NF != 0 {
-    if (NF > max_nf)
-        max_nf = NF
+NF {
+    if (NF > nf_max)
+        nf_max = NF
 
-    ### iterate over columns
-    for (y=1; y<=max_nf; y++) {
-        if (match(dheader[y], $y))
+    ### orig columns
+    for (y=1; y<=nf_max; y++) {
+        if ($y == header[y] || $y == dheader[y])
             continue
         if ($y ~ number)
-            printf("%g", $y)
-        if (y <= max_nf)
-            printf("%s", OFS)
+            printf($y)
+        else
+            printf("")
+        printf(OFS)
     }
 
-    for (y=1; y<=max_nf; y++) {
+    ### diff columns
+    for (y=1; y<=nf_max; y++) {
         if ($y ~ number) {
             data[y] = $y
-            (data_prev[y] ~ number) ?  diff[y] = data[y] - data_prev[y] : diff[y] = "nan"
+            (data_prev[y] ~ number) ? diff[y] = data[y] - data_prev[y] : diff[y] = ""
             data_prev[y] = data[y]
-            printf("%g", diff[y])
-            if (y < max_nf)
-                printf("%s", OFS)
+            printf(diff[y])
+            if (y < nf_max)
+                printf(OFS)
         }
+        diff[y] = ""
     }
     printf("\n")
 }