3.57 mstats - Calculate Statistics of 1 Variable

Specify the numeric fields in the parameter f=, and calculate the statistics specified in the parameter c=. Specify the aggregate key unit at k=. NULL value in the specified field(s) at f= are ignored. However, if all records include NULL values, NULL values will be included in the output.

Format

mstats c= f= [k=] [i=] [o=] [-nfn] [-nfno] [-x] [-q] [precision=] [--help] [--version]

Parameters

`k=`	Compute aggregate statistics on the key field(s) specified (multiple fields can be specified).
`f=`	Fields for which statistics are computed (multiple fields can be specified).
`c=`	Statistics (select one from the list below)
	`sum\|mean\|count\|ucount\|devsq\|var\|uvar\|sd\|usd\|USD\|cv\|min\|qtile1\|`
	`median\|qtile3\|max\|range\|qrange\|mode\|skew\|uskew\|kurt\|ukurt`

List of statistics

Value of `c=`	Description	Equation	Remarks
count	Count (Except NULL value)	$n$ : Number of non-NULL records	It can not be applied to character string field.
ucount	Unique count	$un$ : Number of duplicate values removed	It can not be applied to character string field.
sum	Total	$sum=\sum _{i=1}^ n x_ i$
mean	Arithmetic mean	$m=\frac{1}{n}\sum _{i=1}^ n x_ i$
devsq	Sum of squared deviation	$S=\sum _{i=1}^ n(x_ i-m)^2$
var	Variance	$s^2=\frac{1}{n}S$
uvar	Variance (unbiased estimate)	$u^2=\frac{1}{n-1}S$
sd	Standard deviation	$s=\sqrt {s^2}$
usd	Standard deviation (unbiased variance)	$u=\sqrt {u^2}$	commonly used standard deviation
USD	Unbiased standard deviation	Omission	Accurate unbiased estimation
cv	Coefficient of variation	$cv=s/mx100\%$
mode	Mode	$mode$ : Most frequent value	Print the value of the smaller value if the frequency is same
			Print NULL if values are different.
min	Minimum value	$min=\min _ i x_ i$
max	Maximum value	$max=\max _ i x_ i$
range	Range	$r=max-min$
median	Median	$Q2=Second quartile when sorted in ascending order$
qtile1	First quartile	$Q1=First quartile when sorted in ascending order$
qtile3	Third quartile	$Q3=Third quartile when sorted in ascending order$
qrange	Interquartile range	$rq=Q3-Q1$
skew	Skewness	$\frac{\frac{1}{n}\sum _{i=1}^ n (x_ i-m)^3}{s^3}$
uskew	Skewness (unbiased estimate)	omitted
kurt	Kurtosis	$\frac{\frac{1}{n}\sum _{i=1}^ n (x_ i-m)^4}{s^4}-3.0$
ukurt	Kurtosis (unbiased estimated)	omitted

Examples

Example 1: Basic

Calculate the statistical sum of "quantity" and "amount" field for each "customer".

$ more dat1.csv
customer,quantity,amount
A,1,10
B,5,20
B,2,10
C,1,15
C,3,10
C,1,21
$ mstats k=customer f=quantity,amount c=sum i=dat1.csv o=rsl1.csv
#END# kgstats c=sum f=quantity,amount i=dat1.csv k=customer o=rsl1.csv
$ more rsl1.csv
customer%0,quantity,amount
A,1,10
B,7,30
C,5,46

Example 2: Basic Example 2

Calculate the statistical maximum value.

$ mstats k=customer f=quantity,amount c=max i=dat1.csv o=rsl2.csv
#END# kgstats c=max f=quantity,amount i=dat1.csv k=customer o=rsl2.csv
$ more rsl2.csv
customer%0,quantity,amount
A,1,10
B,5,20
C,3,21

Related Commands

msim : Find out the bivariate statistics.

mavg : Commands specific toc=avg.

msum : Commands specific to c=sum.

mcount : Unlike c=count, this count the number of rows for each aggregate key.