histogram.md

Histograms in Powershell

You can use the group-object to do standard grouping.

This shows all of the files in the current folder, grouped into buckets according to their size. Each bucket is 1000 bytes.

dir |
	group-object -property { [int]($_.length / 1000) } | sort @{e={$_.Name -as [int]}} |
		format-table @{Expression={[string]([int]($_.Name) * 1000) + "-" + [string](-1+([int]($_.Name)+1) * 1000)};Label="Bucket";width=10},
			@{Expression={$_.Count};Label="Count";width=10},
			@{Expression={$_.Group};Label="Members";width=50} -autosize

Consider running the initial dir recursively with -rec

Group by file type

Exclude directories (but include their files)

dir -rec -file | group-object -property extension | sort count -desc

(An older method was to use dir -rec | ?{ $_.PSIsContainer -ne $true } ... instead of -file)

Bucket all files

...for looking at size of all markdown files in TIL, in buckets of 300 bytes each (but changing the bucket size is easy...)

$bucketSize = 300; dir . *.md -rec -file |
group -property { [int]($_.length / $bucketSize) } | sort @{e={$_.Name -as [int]}} |
	format-table @{Expression={[string]([int]($_.Name) * $bucketSize) + "-" + [string](-1+([int]($_.Name)+1) * $bucketSize)};Label="Bucket";width=10},
		@{Expression={$_.Count};Label="Count";width=10},
		@{Expression={$_.Group};Label="Members";width=50} -autosize

---

See also

• SQL Server: Example query for bucketing up a bunch of values