mirror of
https://github.com/johnkerl/miller.git
synced 2026-01-23 10:15:36 +00:00
66 lines
4.6 KiB
HTML
66 lines
4.6 KiB
HTML
POKI_PUT_TOC_HERE
|
|
|
|
<p/>
|
|
<button style="font-weight:bold;color:maroon;border:0" onclick="bodyToggler.expandAll();" href="javascript:;">Expand all sections</button>
|
|
<button style="font-weight:bold;color:maroon;border:0" onclick="bodyToggler.collapseAll();" href="javascript:;">Collapse all sections</button>
|
|
|
|
<h1>flins data</h1>
|
|
<button style="font-weight:bold;color:maroon;border:0" padding=0 onclick="bodyToggler.toggle('body_section_toggle_flins');" href="javascript:;">Toggle section visibility</button>
|
|
<div id="body_section_toggle_flins" style="display: block">
|
|
|
|
<p/> The <a href="data/flins.csv">flins.csv</a> file is some sample data
|
|
obtained from <a href="https://support.spatialkey.com/spatialkey-sample-csv-data">https://support.spatialkey.com/spatialkey-sample-csv-data</a>.
|
|
|
|
<p/>Vertical-tabular format is good for a quick look at CSV data layout — seeing what columns you have to work with:
|
|
POKI_RUN_COMMAND{{head -n 2 data/flins.csv | mlr --icsv --oxtab cat}}HERE
|
|
<p/> A few simple queries:
|
|
POKI_RUN_COMMAND{{mlr --from data/flins.csv --icsv --opprint count-distinct -f county | head}}HERE
|
|
POKI_RUN_COMMAND{{mlr --from data/flins.csv --icsv --opprint count-distinct -f construction,line}}HERE
|
|
<p/> Categorization of total insured value:
|
|
POKI_RUN_COMMAND{{mlr --from data/flins.csv --icsv --opprint stats1 -a min,mean,max -f tiv_2012}}HERE
|
|
POKI_RUN_COMMAND{{mlr --from data/flins.csv --icsv --opprint stats1 -a min,mean,max -f tiv_2012 -g construction,line}}HERE
|
|
|
|
POKI_RUN_COMMAND{{mlr --from data/flins.csv --icsv --oxtab stats1 -a p0,p10,p50,p90,p95,p99,p100 -f hu_site_deductible}}HERE
|
|
POKI_RUN_COMMAND{{mlr --from data/flins.csv --icsv --opprint stats1 -a p95,p99,p100 -f hu_site_deductible -g county then sort -f county | head}}HERE
|
|
POKI_RUN_COMMAND{{mlr --from data/flins.csv --icsv --oxtab stats2 -a corr,linreg-ols,r2 -f tiv_2011,tiv_2012}}HERE
|
|
POKI_RUN_COMMAND{{mlr --from data/flins.csv --icsv --opprint stats2 -a corr,linreg-ols,r2 -f tiv_2011,tiv_2012 -g county}}HERE
|
|
|
|
</div>
|
|
<h1>Color/shape data</h1>
|
|
<button style="font-weight:bold;color:maroon;border:0" padding=0 onclick="bodyToggler.toggle('body_section_toggle_color_shape');" href="javascript:;">Toggle section visibility</button>
|
|
<div id="body_section_toggle_color_shape" style="display: block">
|
|
|
|
<p/> The <a href="data/colored-shapes.dkvp">colored-shapes.dkvp</a> file is some sample data produced by the
|
|
<a href="https://github.com/johnkerl/miller/blob/master/doc/datagen/mkdat2">mkdat2</a> script. The idea is
|
|
<ul>
|
|
<li> Produce some data with known distributions and correlations, and verify that Miller recovers those properties empirically.
|
|
<li> Each record is labeled with one of a few colors and one of a few shapes.
|
|
<li> The <code>flag</code> field is 0 or 1, with probability dependent on color
|
|
<li> The <code>u</code> field is plain uniform on the unit interval.
|
|
<li> The <code>v</code> field is the same, except tightly correlated with <code>u</code> for red circles.
|
|
<li> The <code>w</code> field is autocorrelated for each color/shape pair.
|
|
<li> The <code>x</code> field is boring Gaussian with mean 5 and standard deviation about 1.2, with no dependence on color or shape.
|
|
</ul>
|
|
|
|
<p/> Peek at the data:
|
|
POKI_RUN_COMMAND{{wc -l data/colored-shapes.dkvp}}HERE
|
|
POKI_RUN_COMMAND{{head -n 6 data/colored-shapes.dkvp | mlr --opprint cat}}HERE
|
|
|
|
<p/> Look at uncategorized stats (using <a href="https://github.com/johnkerl/scripts/blob/master/fundam/creach"><code>creach</code></a> for spacing).
|
|
Here it looks reasonable that <code>u</code> is unit-uniform; something’s up with <code>v</code> but we can’t yet see what:
|
|
POKI_RUN_COMMAND{{mlr --oxtab stats1 -a min,mean,max -f flag,u,v data/colored-shapes.dkvp | creach 3}}HERE
|
|
<p/>The histogram shows the different distribution of 0/1 flags:
|
|
POKI_RUN_COMMAND{{mlr --opprint histogram -f flag,u,v --lo -0.1 --hi 1.1 --nbins 12 data/colored-shapes.dkvp}}HERE
|
|
|
|
<p/> Look at univariate stats by color and shape. In particular,
|
|
color-dependent flag probabilities pop out, aligning with their original
|
|
Bernoulli probablities from the data-generator script:
|
|
|
|
POKI_RUN_COMMAND{{mlr --opprint stats1 -a min,mean,max -f flag,u,v -g color then sort -f color data/colored-shapes.dkvp}}HERE
|
|
POKI_RUN_COMMAND{{mlr --opprint stats1 -a min,mean,max -f flag,u,v -g shape then sort -f shape data/colored-shapes.dkvp}}HERE
|
|
|
|
<p/> Look at bivariate stats by color and shape. In particular, <code>u,v</code> pairwise correlation for red circles pops out:
|
|
POKI_RUN_COMMAND{{mlr --opprint --right stats2 -a corr -f u,v,w,x data/colored-shapes.dkvp}}HERE
|
|
POKI_RUN_COMMAND{{mlr --opprint --right stats2 -a corr -f u,v,w,x -g color,shape then sort -nr u_v_corr data/colored-shapes.dkvp}}HERE
|
|
|
|
</div>
|