من الطولي إلى العريض باستخدام pivot_wider()

متى نستخدم التنسيق العريض

library(tidyr)

# Long format: one row per student-quarter combination
long_df <- data.frame(
  student = c('Alice','Alice','Alice','Bob','Bob','Bob'),
  quarter = c('Q1','Q2','Q3','Q1','Q2','Q3'),
  score = c(85, 88, 92, 90, 85, 88)
)

print(long_df)
cat('\n', nrow(long_df), 'rows x', ncol(long_df), 'cols')

أساسيات pivot_wider()

library(tidyr)

long_df <- data.frame(
  student = c('Alice','Alice','Alice','Bob','Bob','Bob'),
  quarter = c('Q1','Q2','Q3','Q1','Q2','Q3'),
  score = c(85, 88, 92, 90, 85, 88)
)

wide_df <- pivot_wider(
  long_df,
  names_from = quarter,
  values_from = score
)

print(wide_df)

id_cols — تحديد مفاتيح الصفوف

library(tidyr)

df <- data.frame(
  year = c(2023,2023,2024,2024),
  region = c('East','West','East','West'),
  quarter = c('Q1','Q1','Q1','Q1'),
  revenue = c(100, 200, 120, 220)
)

pivot_wider(
  df,
  id_cols = c(year, quarter),
  names_from = region,
  values_from = revenue
)

values_fill — معالجة التركيبات المفقودة

library(tidyr)

df <- data.frame(
  product = c('A','A','B'),
  channel = c('online','store','online'),
  sales = c(100, 150, 200)
)

# 'B' has no 'store' data — fill with 0
pivot_wider(
  df,
  names_from = channel,
  values_from = sales,
  values_fill = 0
)

names_prefix — إضافة بادئات

library(tidyr)

df <- data.frame(
  student = c('Alice','Alice','Bob','Bob'),
  quarter = c(1, 2, 1, 2),
  score = c(85, 90, 78, 82)
)

# Without prefix: columns named '1' and '2' (problematic)
# With prefix: q1, q2
pivot_wider(
  df,
  names_from = quarter,
  values_from = score,
  names_prefix = 'q'
)

أعمدة متعددة في values_from

library(tidyr)

df <- data.frame(
  region = c('East','East','West','West'),
  period = c('Q1','Q2','Q1','Q2'),
  revenue = c(100, 120, 200, 210),
  units = c(10, 12, 20, 22)
)

pivot_wider(
  df,
  names_from = period,
  values_from = c(revenue, units)
)

معالجة التكرارات باستخدام values_fn

library(tidyr)

# Duplicate rows: Alice has two Q1 scores
df <- data.frame(
  student = c('Alice','Alice','Alice','Bob','Bob'),
  quarter = c('Q1','Q1','Q2','Q1','Q2'),
  score = c(85, 87, 90, 78, 82)
)

# Resolve duplicates by taking the mean
pivot_wider(
  df,
  names_from = quarter,
  values_from = score,
  values_fn = mean
)

استخدام values_fn مع قائمة مسمّاة

library(tidyr)

df <- data.frame(
  region = c('East','East','West','West'),
  period = c('Q1','Q1','Q1','Q1'),
  revenue = c(100, 120, 200, 210),
  orders = c(10, 12, 20, 22)
)

pivot_wider(
  df,
  names_from = period,
  values_from = c(revenue, orders),
  values_fn = list(revenue = sum, orders = sum)
)

إنشاء جداول التقاطع

library(tidyr)
library(dplyr)

survey <- data.frame(
  dept = c('HR','HR','Eng','Eng','HR','Eng'),
  response = c('Yes','No','Yes','Yes','Yes','No')
)

survey %>%
  count(dept, response) %>%
  pivot_wider(
    names_from = response,
    values_from = n,
    values_fill = 0
  )

التحويل ذهابًا وإيابًا: طويل ← عريض ← طويل

library(tidyr)

original_long <- data.frame(
  id = c(1,1,2,2),
  key = c('a','b','a','b'),
  value = c(10, 20, 30, 40)
)

# Go wide
wide <- pivot_wider(original_long, names_from=key, values_from=value)
print(wide)

# Go long again
pivot_longer(wide, -id, names_to='key', values_to='value')

تطبيقي: تقرير الدرجات

library(tidyr)
library(dplyr)

scores <- data.frame(
  student = c('Alice','Alice','Bob','Bob','Carol','Carol'),
  subject = c('Math','Science','Math','Science','Math','Science'),
  score = c(92, 88, 85, 91, 79, 83)
)

scores %>%
  pivot_wider(names_from=subject, values_from=score) %>%
  mutate(average = round((Math + Science) / 2, 1)) %>%
  arrange(desc(average))

تحقق سريع

مراجعة: pivot_wider()

library(tidyr)
library(dplyr)

data.frame(
  region = c('East','East','West','West'),
  metric = c('revenue','units','revenue','units'),
  value = c(100, 10, 200, 20)
) %>%
  pivot_wider(
    names_from = metric,
    values_from = value,
    values_fill = 0
  ) %>%
  mutate(revenue_per_unit = revenue / units)