From ddcb1f73423a9b47d9b457cf3b19f622b0aa567f Mon Sep 17 00:00:00 2001
From: wukong <wukong@longaeva>
Date: Sat, 5 Jan 2019 16:33:14 -0800
Subject: added min and max to mean and mean_avg, fixed regex bugs in mean

---
 kalman_init.awk | 10 +++++-----
 mean.awk        | 29 ++++++++++++++---------------
 mean_avg.awk    | 10 ++++++----
 3 files changed, 25 insertions(+), 24 deletions(-)

diff --git a/kalman_init.awk b/kalman_init.awk
index 38cb66c..57f7bbd 100644
--- a/kalman_init.awk
+++ b/kalman_init.awk
@@ -54,22 +54,22 @@ NF != 0 {
 
             ### kalman gain
             if (count[n] <= 1) {
-                KG[n] = 1
+                Gk[n] = 1
                 est_err[n] = sqrt(diff[n]*diff[n]) + 2.0*sqrt(diff2[n]*diff2[n])
                 est_err_last[n] = sqrt(diff[n]*diff[n]) + 2.0*sqrt(diff2[n]*diff2[n])
             }
             else
-                KG[n] = est_err_last[n]/(est_err_last[n] + meas_err[n])
+                Gk[n] = est_err_last[n]/(est_err_last[n] + meas_err[n])
 
             ### update estimate and estimated error
-            est[n] = est_last[n] + KG[n]*($n - est_last[n])
-            est_err[n] = (1.0 - KG[n])*est_err_last[n]
+            est[n] = est_last[n] + Gk[n]*($n - est_last[n])
+            est_err[n] = (1.0 - Gk[n])*est_err_last[n]
 
             ### visual check
             printf "meas: " OFS OFMT OFS OFMT OFS OFMT, $n, meas_err[n], count[n]
             printf "d1: " OFS OFMT, diff[n]
             printf "d2: " OFS OFMT, diff2[n]
-            printf "est: " OFS OFMT OFS OFMT OFS OFMT, est[n], est_err[n], KG[n]
+            printf "est: " OFS OFMT OFS OFMT OFS OFMT, est[n], est_err[n], Gk[n]
 
             ### update previously remembered values
             last2[n] = last[n]
diff --git a/mean.awk b/mean.awk
index 816208c..b6a794e 100644
--- a/mean.awk
+++ b/mean.awk
@@ -1,36 +1,35 @@
 #!/usr/bin/awk -f
 
 ### mean.awk
-# calculate mean average
+# calculate mean average of serialized numbers
 
 BEGIN {
     OFS = FS
-    sign = "[+-±]?"
+    sign = "[+-]?"
     decimal = "[0-9]+[.]?[0-9]*"
-    fraction = "[.][0-9]*"
+    fraction = "[.][0-9]+"
     exponent = "([Ee]" sign "[0-9]+)?"
     number = "^" sign "(" decimal "|" fraction ")" exponent "$"
 }
 
 # Welford's 'online' algorithm for variance
 NF > 0 {
-    for (m=1; m<=NF; m++) {
-        if (m ~ number) {
+    for (n=1; n<=NF; n++) {
+        if ($n ~ number) {
             count += 1
-            delta0 = $m - mean
-            mean = mean + delta0/count
-            delta1 = $m - mean
-            M2 = M2 + delta0*delta1
-            if (count > 1)
-                var = M2/(count - 1)
-            else
-                var = "nan"
+            (count == 1 || $n < min) ? min = $n : min = min
+            (count == 1 || $n > max) ? max = $n : max = max
+            delta0 = $n - mean
+            mean += delta0/count
+            delta1 = $n - mean
+            M2 += delta0*delta1
+            (count > 1) ? var = M2/(count - 1) : var = "nan"
         }
     }
 }
 
 END {
-    print "mean", "std_err", "std_dev", "count"
-    print mean, sqrt(var/count), sqrt(var), count
+    print "mean", "std_err", "std_dev", "min", "max", "count"
+    print mean, sqrt(var/count), sqrt(var), min, max, count
 }
 
diff --git a/mean_avg.awk b/mean_avg.awk
index ed6f81a..4bf7051 100644
--- a/mean_avg.awk
+++ b/mean_avg.awk
@@ -24,22 +24,24 @@ NF > 0 {
     for (n=1; n <= NF; n++) {
         if ($n ~ number) {
             count[n] += 1
+            (count[n] == 1 || $n < min[n]) ? min[n] = $n : min[n] = min[n]
+            (count[n] == 1 || $n > max[n]) ? max[n] = $n : max[n] = max[n]
             delta0[n] = $n - mean[n]
             mean[n] += delta0[n]/count[n]
             delta1[n] = $n - mean[n]
             M2[n] += delta0[n]*delta1[n]
-            (count[n] > 1) ? var[n] = M2[n]/(count[n] - 1) : var[n] = ""
+            (count[n] > 1) ? var[n] = M2[n]/(count[n] - 1) : var[n] = "nan"
         }
     }
 }
 
 END {
-    print "col", "mean", "std_err", "std_dev", "count"
+    print "col", "mean", "std_err", "std_dev", "min", "max", "count"
     for (n=1; n<=max_nf; n++) {
         if (header[n])
-            print header[n], mean[n], sqrt(var[n]/count[n]), sqrt(var[n]), count[n]
+            print header[n], mean[n], sqrt(var[n]/count[n]), sqrt(var[n]), min[n], max[n], count[n]
         else
-            print "col" n, mean[n], sqrt(var[n]/count[n]), sqrt(var[n]), count[n]
+            print "col" n, mean[n], sqrt(var[n]/count[n]), sqrt(var[n]), min[n], max[n], count[n]
     }
 }
 
-- 
cgit v1.2.3