Ing. de Sistemas, Magister Ing. Industrial. #DataAnalysis #DataScience #MachineLearning #Football
output = tidy(
cars,
groupBy(
['cyl', 'gear'],
[
summarize({
n: n(),
mpg: mean('mpg'),
hp: mean('hp'),
wt: mean('wt')
})
]
),
select(['cyl', 'gear', everything()]),
arrange([desc('n'), desc('mpg')])
)
table(
tidy(
table1,
mutate({
rate: d => (d.cases / d.population) * 10000,
'|': '|', // visual spacer for table()
// alternatively, use the `rate()` function for safety around division by 0 and nully values
// rate2: rate('cases', 'population'),
rate2: rate(d => d.cases * 10000, 'population'),
// or the TMath.rate function directly:
rate3: d => TMath.rate(d.cases, d.population) * 10000
})
)
)
table(tidy(table1, count('year', { wt: 'cases' })))
table(table4a)
table(
tidy(
table4a,
pivotLonger({
cols: ['1999', '2000'],
namesTo: 'year',
valuesTo: 'cases'
})
)
)
table(
tidy(
table4b,
pivotLonger({
cols: ['1999', '2000'],
namesTo: 'year',
valuesTo: 'population'
})
)
)
{
let tidy4a = tidy(
table4a,
pivotLonger({
cols: ['1999', '2000'],
namesTo: 'year',
valuesTo: 'cases'
})
);
let tidy4b = tidy(
table4b,
pivotLonger({
cols: ['1999', '2000'],
namesTo: 'year',
valuesTo: 'population'
})
);
return table(tidy(tidy4a, leftJoin(tidy4b)));
}
table(table2)
table(tidy(table2, pivotWider({ namesFrom: 'type', valuesFrom: 'count' })))
// or re-arrange
table(
tidy(
table2,
pivotWider({ namesFrom: 'type', valuesFrom: 'count' }),
select(['country', 'year', everything()])
)
)
table(table3)
// tidy doesn't provide a separate command (yet?), but you could use mutate
table(
tidy(
table3,
mutate({
rate: d => d.rate.split('/'),
cases: d => +d.rate[0],
population: d => +d.rate[1]
}),
select('-rate')
)
)
// or you could use map
table(
tidy(
table3,
map(d => {
const [cases, population] = d.rate.split('/').map(x => +x);
return { ...d, cases, population };
}),
select('-rate')
)
)
// there's no unite (yet), but a mutate + select will do the trick
table(
tidy(
table5,
mutate({ new: d => `${d.century}_${d.year}` }),
select(['-century', '-year'])
)
)
table(
tidy(
table5,
mutate({ new: d => +`${d.century}${d.year}` }),
select(['-century', '-year'])
)
)
table(stocks)
table(tidy(stocks, pivotWider({ namesFrom: 'year', valuesFrom: 'return' })))
// drop out missing values
table(
tidy(
stocks,
pivotWider({ namesFrom: 'year', valuesFrom: 'return' }),
pivotLonger({
cols: ['2015', '2016'],
namesTo: 'year',
valuesTo: 'return'
}),
// pivotLonger does not have values_drop_na at the moment
filter(d => d.return != null)
)
)
// use complete to make missing values explicit
table(tidy(stocks, complete(['year', 'qtr'])))
treatment = [
{ person: 'Derrick Whitmore', treatment: 1, response: 7 },
{ treatment: 2, response: 10 },
{ person: undefined, treatment: 3, response: 9 },
{ person: 'Katherine Burke', treatment: 1, response: 4 }
]
table(treatment)
// fill in columns with last seen values
table(tidy(treatment, fill('person')))
table(who.slice(0, 20))
who1 = tidy(
who,
pivotLonger({ cols: startsWith('new'), namesTo: 'key', valuesTo: 'cases' }),
filter(d => d.cases !== 'NA')
)
table(who1.slice(0, 20))
table(tidy(who1, count('key')))
who2 = tidy(who1, mutate({ key: d => d.key.replace('newrel', 'new_rel') }))
table(who2.filter(d => d.key.includes('new_rel')).slice(0, 15))
who3 = tidy(
who2,
mutate({
key: d => d.key.split('_'),
new: d => d.key[0],
type: d => d.key[1],
sexage: d => d.key[2]
}),
select(['-key'])
)
table(who3.slice(0, 6))
table(tidy(who3, count('new')))
who4 = tidy(who3, select(['-new', '-iso2', '-iso3']))
who5 = tidy(
who4,
mutate({
sexage: d => [d.sexage[0], d.sexage.substring(1)],
sex: d => d.sexage[0],
age: d => +d.sexage[1]
}),
select('-sexage')
)
table(who5.slice(0, 10))
// all together now
table(
tidy(
who,
pivotLonger({ cols: startsWith('new'), namesTo: 'key', valuesTo: 'cases' }),
filter(d => d.cases !== 'NA'),
mutate({ key: d => d.key.replace('newrel', 'new_rel') }),
// separate equivalent:
mutate({
key: d => d.key.split('_'),
new: d => d.key[0],
type: d => d.key[1],
sexage: d => d.key[2]
}),
// separate equivalent:
mutate({
sexage: d => [d.sexage[0], d.sexage.substring(1)],
sex: d => d.sexage[0],
age: d => +d.sexage[1]
}),
select(['-key', '-new', '-iso2', '-iso3', '-sexage'])
).slice(0, 20)
)
viewof tidyjsfile = html`<input type=file accept=".js">`
d3 = require('d3-array', 'd3-dsv')
module = require(URL.createObjectURL(tidyjsfile))
import {
tidy,
startsWith,
groupBy,
summarize,
sum,
mutate,
select,
n,
mean,
everything,
arrange,
desc,
filter,
rate,
TMath,
pivotWider,
pivotLonger,
leftJoin,
count,
map,
complete,
fill
} from '@pbeshai/tidyjs' // observable import
module_attachment = require(await FileAttachment("tidy.js").url())
import {Table} from "@observablehq/inputs"
// switch from tmcw table to observablehq and just use same name
table = Table
table1csv = `country,year,cases,population
Afghanistan,1999,745,19987071
Afghanistan,2000,2666,20595360
Brazil,1999,37737,172006362
Brazil,2000,80488,174504898
China,1999,212258,1272915272
China,2000,213766,1280428583`
table1 = d3.csvParse(table1csv, d => ({ ...d, year: +d.year, cases: +d.cases, population: +d.population }))
table2csv = `country,year,type,count
Afghanistan,1999,cases,745
Afghanistan,1999,population,19987071
Afghanistan,2000,cases,2666
Afghanistan,2000,population,20595360
Brazil,1999,cases,37737
Brazil,1999,population,172006362
Brazil,2000,cases,80488
Brazil,2000,population,174504898
China,1999,cases,212258
China,1999,population,1272915272
China,2000,cases,213766
China,2000,population,1280428583`
table2 = d3.csvParse(table2csv, d => ({ ...d, year: +d.year, count: +d.count }))
table3csv=`country,year,rate
Afghanistan,1999,745/19987071
Afghanistan,2000,2666/20595360
Brazil,1999,37737/172006362
Brazil,2000,80488/174504898
China,1999,212258/1272915272
China,2000,213766/1280428583
`
table3 = d3.csvParse(table3csv, d => ({ ...d, year: +d.year }))
table4acsv = `country,1999,2000
Afghanistan,745,2666
Brazil,37737,80488
China,212258,213766`
table4a = d3.csvParse(table4acsv, d => ({ ...d, '1999': +d['1999'], '2000': +d['2000'] }))
table4bcsv = `country,1999,2000
Afghanistan,19987071,20595360
Brazil,172006362,174504898
China,1272915272,1280428583`
table4b = d3.csvParse(table4bcsv, d => ({ ...d, '1999': +d['1999'], '2000': +d['2000'] }))
table5 = tidy(
table3,
mutate({
century: d => `${Math.floor(d.year / 100)}`,
year: d => `${d.year % 100}`.padStart(2, '0')
}),
select(['country', 'century', 'year', 'rate'])
)
whocsv = await FileAttachment("who.csv").text()
who = d3.csvParse(whocsv, d => ({ ...d, year: +d.year,}))
stocks = [
{ year: 2015, qtr: 1, return: 1.88 },
{ year: 2015, qtr: 2, return: 0.59 },
{ year: 2015, qtr: 3, return: 0.35 },
{ year: 2015, qtr: 4, return: null },
{ year: 2016, qtr: 2, return: 0.92 },
{ year: 2016, qtr: 3, return: 0.17 },
{ year: 2016, qtr: 4, return: 2.66 }
]
Purpose-built for displays of data
Observable is your go-to platform for exploring data and creating expressive data visualizations. Use reactive JavaScript notebooks for prototyping and a collaborative canvas for visual data exploration and dashboard creation.
