From ce5e0a8f088e19e8ebe9a27d8efd4207363791af Mon Sep 17 00:00:00 2001
From: wukong <wukong@longaeva>
Date: Fri, 15 Jun 2018 09:17:23 -0700
Subject: fixed row 1 header issue in diff.awk; other small adjustments in
 hamming and lin_reg

---
 conv.awk        |  1 +
 diff.awk        | 46 ++++++++++++++++++++++++++--------------------
 diff1.awk       |  1 +
 hamming.awk     |  6 +++---
 kalman.awk      |  1 +
 kalman_init.awk |  1 +
 lin_reg.awk     |  2 +-
 lin_reg1.awk    |  1 +
 8 files changed, 35 insertions(+), 24 deletions(-)

diff --git a/conv.awk b/conv.awk
index 2248927..b1e9e62 100644
--- a/conv.awk
+++ b/conv.awk
@@ -4,6 +4,7 @@
 # [PoC] linear convolution (with hardcoded IR window).
 
 BEGIN {
+    OFMT = "%.9g"
     X = ARGV[1]
     input_size = split(X, X_arr)
     #H = "1.0 1.0 1.0"      # rect
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")
 }
diff --git a/diff1.awk b/diff1.awk
index cb86e00..42cc5a8 100644
--- a/diff1.awk
+++ b/diff1.awk
@@ -4,6 +4,7 @@
 # numerical diff along columns
 
 BEGIN {
+    OFMT = "%.9g"
     sign = "[+-]?"
     decimal = "[0-9]+[.]?[0-9]*"
     fraction = "[.][0-9]*"
diff --git a/hamming.awk b/hamming.awk
index 21d241e..84dacd3 100644
--- a/hamming.awk
+++ b/hamming.awk
@@ -6,10 +6,10 @@
 # 'a0' and 'a1' parameters of the raised cosine.
 
 BEGIN {
-
+    OMFT = "%.9g"
     N = ARGV[1]
-
     pi = 4*(4*atan2(1,5) - atan2(1,239))
+
     #a0 = (25.0/46.0)
     #a1 = (21.0/46.0)
 
@@ -25,7 +25,7 @@ BEGIN {
             w[n] = a0 - a1*cos((2*pi*n)/(N - 1))
             sum_w += w[n]
         }
-        printf("%g  %g\n", w[n], 2.0*(w[n])/(N + 1))
+        print 2.0*(w[n])/(N + 1)
     }
 
 }
diff --git a/kalman.awk b/kalman.awk
index f67e9de..41936e2 100644
--- a/kalman.awk
+++ b/kalman.awk
@@ -4,6 +4,7 @@
 # simple linear regression between columns
 
 BEGIN {
+    OMFT = "%.9g"
     sign = "[+-]?"
     decimal = "[0-9]+[.]?[0-9]*"
     fraction = "[.][0-9]*"
diff --git a/kalman_init.awk b/kalman_init.awk
index f1dcac4..9e0cd14 100644
--- a/kalman_init.awk
+++ b/kalman_init.awk
@@ -4,6 +4,7 @@
 # kalman csv experiemnt
 
 BEGIN {
+    OMFT = "%.9"
     sign = "[+-]?"
     decimal = "[0-9]+[.]?[0-9]*"
     fraction = "[.][0-9]*"
diff --git a/lin_reg.awk b/lin_reg.awk
index e1cb722..51f2e87 100644
--- a/lin_reg.awk
+++ b/lin_reg.awk
@@ -39,7 +39,7 @@ NF != 0 {
             sum_delta2[y] += delta[y]*delta[y]
 
             ### sample variance
-            (count[y] - 1) ? var[y] = sum_delta2[y]/(count[y] - 1) : var[y] = 0
+            (count[y] > 1) ? var[y] = sum_delta2[y]/(count[y] - 1) : var[y] = 0
 
             # x = row, y = col
             for (x=1; x<=max_nf; x++) {
diff --git a/lin_reg1.awk b/lin_reg1.awk
index de1189f..cd3db8e 100644
--- a/lin_reg1.awk
+++ b/lin_reg1.awk
@@ -4,6 +4,7 @@
 # simple linear regression between individual text columns
 
 BEGIN {
+    OFS = "%.9g"
     sign = "[+-]?"
     decimal = "[0-9]+[.]?[0-9]*"
     fraction = "[.][0-9]*"
-- 
cgit v1.2.3