diff options
Diffstat (limited to 'mean_avg.awk')
| -rw-r--r-- | mean_avg.awk | 9 |
1 files changed, 5 insertions, 4 deletions
diff --git a/mean_avg.awk b/mean_avg.awk index e4596b0..bef984d 100644 --- a/mean_avg.awk +++ b/mean_avg.awk @@ -20,7 +20,7 @@ NR == 1 { } # Welford's 'online' algorithm for variance -NF > 0 { +NF { if (NF > max_nf) max_nf = NF for (n=1; n <= NF; n++) { @@ -28,6 +28,7 @@ NF > 0 { 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] + range[n] = max[n] - min[n] delta0[n] = $n - mean[n] mean[n] += delta0[n]/count[n] delta1[n] = $n - mean[n] @@ -38,12 +39,12 @@ NF > 0 { } END { - print "col", "mean", "std_err", "std_dev", "min", "max", "total", "count" + print "col", "mean", "std_err", "std_dev", "range", "min", "max", "total", "count" for (n=1; n<=max_nf; n++) { if (header[n]) - print header[n], mean[n], sqrt(var[n]/count[n]), sqrt(var[n]), min[n], max[n], mean[n]*count[n], count[n] + print header[n], mean[n], sqrt(var[n]/count[n]), sqrt(var[n]), range[n], min[n], max[n], mean[n]*count[n], count[n] else - print "col" n, mean[n], sqrt(var[n]/count[n]), sqrt(var[n]), min[n], max[n], mean[n]*count[n], count[n] + print "col" n, mean[n], sqrt(var[n]/count[n]), sqrt(var[n]), range[n], min[n], max[n], mean[n]*count[n], count[n] } } |