INTORDUCTION OF panelr()
Yucheng Zhang
2 DATA PREPARATION
2.1 Making and check panel data format
2.1.1 Integrating panel_data() and dplyr()
## ──────────────────────────────────────────────────────────────────────────────────────────
## exp wks occ ind south smsa ms fem union ed blk lwage t id
## ──────────────────────────────────────────────────────────────────────────────────────────
## 1 3.000 32.000 0.000 0.000 1.000 0.000 1.000 0.000 0.000 9.000 0.000 5.561 1.000 1.000
## 2 4.000 43.000 0.000 0.000 1.000 0.000 1.000 0.000 0.000 9.000 0.000 5.720 2.000 1.000
## 3 5.000 40.000 0.000 0.000 1.000 0.000 1.000 0.000 0.000 9.000 0.000 5.996 3.000 1.000
## 4 6.000 39.000 0.000 0.000 1.000 0.000 1.000 0.000 0.000 9.000 0.000 5.996 4.000 1.000
## 5 7.000 42.000 0.000 1.000 1.000 0.000 1.000 0.000 0.000 9.000 0.000 6.061 5.000 1.000
## 6 8.000 35.000 0.000 1.000 1.000 0.000 1.000 0.000 0.000 9.000 0.000 6.174 6.000 1.000
## 7 9.000 32.000 0.000 1.000 1.000 0.000 1.000 0.000 0.000 9.000 0.000 6.244 7.000 1.000
## 8 30.000 34.000 1.000 0.000 0.000 0.000 1.000 0.000 0.000 11.000 0.000 6.163 1.000 2.000
## 9 31.000 27.000 1.000 0.000 0.000 0.000 1.000 0.000 0.000 11.000 0.000 6.215 2.000 2.000
## 10 32.000 33.000 1.000 1.000 0.000 0.000 1.000 0.000 1.000 11.000 0.000 6.263 3.000 2.000
## 11 33.000 30.000 1.000 1.000 0.000 0.000 1.000 0.000 0.000 11.000 0.000 6.544 4.000 2.000
## 12 34.000 30.000 1.000 1.000 0.000 0.000 1.000 0.000 0.000 11.000 0.000 6.697 5.000 2.000
## 13 35.000 37.000 1.000 1.000 0.000 0.000 1.000 0.000 0.000 11.000 0.000 6.791 6.000 2.000
## 14 36.000 30.000 1.000 1.000 0.000 0.000 1.000 0.000 0.000 11.000 0.000 6.816 7.000 2.000
## 15 6.000 50.000 1.000 1.000 0.000 0.000 1.000 0.000 1.000 12.000 0.000 5.652 1.000 3.000
## 16 7.000 51.000 1.000 1.000 0.000 0.000 1.000 0.000 1.000 12.000 0.000 6.436 2.000 3.000
## 17 8.000 50.000 1.000 1.000 0.000 0.000 1.000 0.000 1.000 12.000 0.000 6.548 3.000 3.000
## 18 9.000 52.000 1.000 1.000 0.000 0.000 1.000 0.000 1.000 12.000 0.000 6.603 4.000 3.000
## 19 10.000 52.000 1.000 1.000 0.000 0.000 1.000 0.000 1.000 12.000 0.000 6.696 5.000 3.000
## 20 11.000 52.000 1.000 1.000 0.000 0.000 0.000 0.000 1.000 12.000 0.000 6.779 6.000 3.000
## ──────────────────────────────────────────────────────────────────────────────────────────# Create `panel_data` object
wages <- panel_data(WageData, id = id, wave = t) %>%
# Pass to mutate, which will calculate statistics groupwise when appropriate
mutate(
wage = exp(lwage), # reverse transform the log wage variable
mean_wage_individual = mean(wage), # means calculated separately by entity
lag_wage = lag(wage) # mutate() will calculate lagged values correctly
) %>%
# Use `panelr`'s complete_data() to filter for entities that have
# enough observations
complete_data(wage, union, min.waves = 5) %>% # drop if there aren't 5 completions
# You can use unpanel() if you need to do rowwise or columnwise operations
unpanel() %>%
mutate(
mean_wage_grand = mean(wage)
) %>%
# You'll need to convert back to panel_data if you want to keep using panelr functions
panel_data(id = id, wave = t)
Describe(wages)## Descriptive Statistics:
## ───────────────────────────────────────────────────────────────────────────────────
## N Mean SD | Median Min Max Skewness Kurtosis
## ───────────────────────────────────────────────────────────────────────────────────
## id** 4165 298.00 171.78 | 298.00 1.00 595.00 0.00 -1.20
## t* 4165 4.00 2.00 | 4.00 1.00 7.00 0.00 -1.25
## exp* 4165 19.85 10.97 | 18.00 1.00 51.00 0.40 -0.93
## wks* 4165 46.81 5.13 | 48.00 5.00 52.00 -2.89 11.91
## occ* 4165 0.51 0.50 | 1.00 0.00 1.00 -0.04 -2.00
## ind* 4165 0.40 0.49 | 0.00 0.00 1.00 0.43 -1.82
## south* 4165 0.29 0.45 | 0.00 0.00 1.00 0.92 -1.15
## smsa* 4165 0.65 0.48 | 1.00 0.00 1.00 -0.65 -1.58
## ms* 4165 0.81 0.39 | 1.00 0.00 1.00 -1.62 0.61
## fem* 4165 0.11 0.32 | 0.00 0.00 1.00 2.45 4.00
## union* 4165 0.36 0.48 | 0.00 0.00 1.00 0.57 -1.68
## ed* 4165 12.85 2.79 | 12.00 4.00 17.00 -0.26 -0.29
## blk* 4165 0.07 0.26 | 0.00 0.00 1.00 3.30 8.91
## lwage* 4165 6.68 0.46 | 6.68 4.61 8.54 -0.04 0.52
## wage* 4165 882.93 442.81 | 800.00 100.00 5100.00 2.13 9.24
## mean_wage_individual* 4165 882.90 0.00 | 882.90 882.90 882.90 NaN NaN
## lag_wage* 4165 882.93 442.81 | 800.00 100.00 5100.00 2.13 9.24
## mean_wage_grand* 4165 882.90 0.00 | 882.90 882.90 882.90 NaN NaN
## ───────────────────────────────────────────────────────────────────────────────────
##
## NOTE: `id` transformed to numeric.## ──────────────────────────────────────────────────────────────────
## id t wage mean_wage_individual lag_wage mean_wage_grand
## ──────────────────────────────────────────────────────────────────
## 1 1 1.000 260.000 882.931 260.000 882.931
## 2 1 2.000 305.000 882.931 305.000 882.931
## 3 1 3.000 401.999 882.931 401.999 882.931
## 4 1 4.000 401.999 882.931 401.999 882.931
## 5 1 5.000 429.001 882.931 429.001 882.931
## 6 1 6.000 480.002 882.931 480.002 882.931
## 7 1 7.000 515.002 882.931 515.002 882.931
## 8 2 1.000 474.998 882.931 474.998 882.931
## 9 2 2.000 500.001 882.931 500.001 882.931
## 10 2 3.000 525.001 882.931 525.001 882.931
## 11 2 4.000 694.999 882.931 694.999 882.931
## 12 2 5.000 809.997 882.931 809.997 882.931
## 13 2 6.000 889.999 882.931 889.999 882.931
## 14 2 7.000 912.000 882.931 912.000 882.931
## 15 3 1.000 285.000 882.931 285.000 882.931
## 16 3 2.000 624.000 882.931 624.000 882.931
## 17 3 3.000 698.001 882.931 698.001 882.931
## 18 3 4.000 737.002 882.931 737.002 882.931
## 19 3 5.000 809.001 882.931 809.001 882.931
## 20 3 6.000 878.996 882.931 878.996 882.931
## ──────────────────────────────────────────────────────────────────2.1.2 complete_data(): Filter out entities with too few observations
#wages <- panel_data(WageData, id = id, wave = t)
Cdata=complete_data(wages, wks, lwage, min.waves = 3)
head(Cdata[,c(1:5)],20)%>%print_table()## ────────────────────────────────
## id t exp wks occ
## ────────────────────────────────
## 1 1 1.000 3.000 32.000 0.000
## 2 1 2.000 4.000 43.000 0.000
## 3 1 3.000 5.000 40.000 0.000
## 4 1 4.000 6.000 39.000 0.000
## 5 1 5.000 7.000 42.000 0.000
## 6 1 6.000 8.000 35.000 0.000
## 7 1 7.000 9.000 32.000 0.000
## 8 2 1.000 30.000 34.000 1.000
## 9 2 2.000 31.000 27.000 1.000
## 10 2 3.000 32.000 33.000 1.000
## 11 2 4.000 33.000 30.000 1.000
## 12 2 5.000 34.000 30.000 1.000
## 13 2 6.000 35.000 37.000 1.000
## 14 2 7.000 36.000 30.000 1.000
## 15 3 1.000 6.000 50.000 1.000
## 16 3 2.000 7.000 51.000 1.000
## 17 3 3.000 8.000 50.000 1.000
## 18 3 4.000 9.000 52.000 1.000
## 19 3 5.000 10.000 52.000 1.000
## 20 3 6.000 11.000 52.000 1.000
## ────────────────────────────────2.1.3 model_frame(): Make dataframe using lm() format
## ──────────────────────────────────────────
## id t lwage lag(union) wks blk
## ──────────────────────────────────────────
## 1 1 1.000 5.561 32.000 0.000
## 2 1 2.000 5.720 0.000 43.000 0.000
## 3 1 3.000 5.996 0.000 40.000 0.000
## 4 1 4.000 5.996 0.000 39.000 0.000
## 5 1 5.000 6.061 0.000 42.000 0.000
## 6 1 6.000 6.174 0.000 35.000 0.000
## 7 1 7.000 6.244 0.000 32.000 0.000
## 8 2 1.000 6.163 34.000 0.000
## 9 2 2.000 6.215 0.000 27.000 0.000
## 10 2 3.000 6.263 0.000 33.000 0.000
## 11 2 4.000 6.544 1.000 30.000 0.000
## 12 2 5.000 6.697 0.000 30.000 0.000
## 13 2 6.000 6.791 0.000 37.000 0.000
## 14 2 7.000 6.816 0.000 30.000 0.000
## 15 3 1.000 5.652 50.000 0.000
## 16 3 2.000 6.436 1.000 51.000 0.000
## 17 3 3.000 6.548 1.000 50.000 0.000
## 18 3 4.000 6.603 1.000 52.000 0.000
## 19 3 5.000 6.696 1.000 52.000 0.000
## 20 3 6.000 6.779 1.000 52.000 0.000
## ──────────────────────────────────────────## Descriptive Statistics:
## ───────────────────────────────────────────────────────────────────────────
## N (NA) Mean SD | Median Min Max Skewness Kurtosis
## ───────────────────────────────────────────────────────────────────────────
## id** 4165 298.00 171.78 | 298.00 1.00 595.00 0.00 -1.20
## t* 4165 4.00 2.00 | 4.00 1.00 7.00 0.00 -1.25
## lwage* 4165 6.68 0.46 | 6.68 4.61 8.54 -0.04 0.52
## lag(union)* 3570 595 0.36 0.48 | 0.00 0.00 1.00 0.57 -1.68
## wks* 4165 46.81 5.13 | 48.00 5.00 52.00 -2.89 11.91
## blk* 4165 0.07 0.26 | 0.00 0.00 1.00 3.30 8.91
## ───────────────────────────────────────────────────────────────────────────
##
## NOTE: `id` transformed to numeric.## Descriptive Statistics:
## ───────────────────────────────────────────────────────────────────────────────────
## N Mean SD | Median Min Max Skewness Kurtosis
## ───────────────────────────────────────────────────────────────────────────────────
## id** 4165 298.00 171.78 | 298.00 1.00 595.00 0.00 -1.20
## t* 4165 4.00 2.00 | 4.00 1.00 7.00 0.00 -1.25
## exp* 4165 19.85 10.97 | 18.00 1.00 51.00 0.40 -0.93
## wks* 4165 46.81 5.13 | 48.00 5.00 52.00 -2.89 11.91
## occ* 4165 0.51 0.50 | 1.00 0.00 1.00 -0.04 -2.00
## ind* 4165 0.40 0.49 | 0.00 0.00 1.00 0.43 -1.82
## south* 4165 0.29 0.45 | 0.00 0.00 1.00 0.92 -1.15
## smsa* 4165 0.65 0.48 | 1.00 0.00 1.00 -0.65 -1.58
## ms* 4165 0.81 0.39 | 1.00 0.00 1.00 -1.62 0.61
## fem* 4165 0.11 0.32 | 0.00 0.00 1.00 2.45 4.00
## union* 4165 0.36 0.48 | 0.00 0.00 1.00 0.57 -1.68
## ed* 4165 12.85 2.79 | 12.00 4.00 17.00 -0.26 -0.29
## blk* 4165 0.07 0.26 | 0.00 0.00 1.00 3.30 8.91
## lwage* 4165 6.68 0.46 | 6.68 4.61 8.54 -0.04 0.52
## wage* 4165 882.93 442.81 | 800.00 100.00 5100.00 2.13 9.24
## mean_wage_individual* 4165 882.90 0.00 | 882.90 882.90 882.90 NaN NaN
## lag_wage* 4165 882.93 442.81 | 800.00 100.00 5100.00 2.13 9.24
## mean_wage_grand* 4165 882.90 0.00 | 882.90 882.90 882.90 NaN NaN
## ───────────────────────────────────────────────────────────────────────────────────
##
## NOTE: `id` transformed to numeric.2.1.4 make_wb_data(): Prepare data for within-between modeling
wb.data=make_wb_data(lwage ~ lag(union) + wks | blk | blk * wks + (lag(union) | id), data = wages)
head(wb.data,20)%>%print_table()## ──────────────────────────────────────────────────────────────────────────────────────────────
## id t lwage lag(union) wks blk imean(lag(union)) imean(wks) imean(wks:blk) wks:blk
## ──────────────────────────────────────────────────────────────────────────────────────────────
## 1 1 2.000 5.720 0.000 5.429 0.000 0.000 37.571 0.000 0.000
## 2 1 3.000 5.996 0.000 2.429 0.000 0.000 37.571 0.000 0.000
## 3 1 4.000 5.996 0.000 1.429 0.000 0.000 37.571 0.000 0.000
## 4 1 5.000 6.061 0.000 4.429 0.000 0.000 37.571 0.000 0.000
## 5 1 6.000 6.174 0.000 -2.571 0.000 0.000 37.571 0.000 -0.000
## 6 1 7.000 6.244 0.000 -5.571 0.000 0.000 37.571 0.000 -0.000
## 7 2 2.000 6.215 -0.167 -4.571 0.000 0.167 31.571 0.000 -0.000
## 8 2 3.000 6.263 -0.167 1.429 0.000 0.167 31.571 0.000 0.000
## 9 2 4.000 6.544 0.833 -1.571 0.000 0.167 31.571 0.000 -0.000
## 10 2 5.000 6.697 -0.167 -1.571 0.000 0.167 31.571 0.000 -0.000
## 11 2 6.000 6.791 -0.167 5.429 0.000 0.167 31.571 0.000 0.000
## 12 2 7.000 6.816 -0.167 -1.571 0.000 0.167 31.571 0.000 -0.000
## 13 3 2.000 6.436 0.000 0.571 0.000 1.000 50.429 0.000 0.000
## 14 3 3.000 6.548 0.000 -0.429 0.000 1.000 50.429 0.000 -0.000
## 15 3 4.000 6.603 0.000 1.571 0.000 1.000 50.429 0.000 0.000
## 16 3 5.000 6.696 0.000 1.571 0.000 1.000 50.429 0.000 0.000
## 17 3 6.000 6.779 0.000 1.571 0.000 1.000 50.429 0.000 0.000
## 18 3 7.000 6.861 0.000 -4.429 0.000 1.000 50.429 0.000 -0.000
## 19 4 2.000 6.238 0.000 -1.857 1.000 0.000 47.857 0.000 -1.857
## 20 4 3.000 6.301 0.000 -1.857 1.000 0.000 47.857 0.000 -1.857
## ──────────────────────────────────────────────────────────────────────────────────────────────2.1.5 make_diff_data(): Generate differenced and asymmetric effects data
## ────────────────────────────────
## id t wks lwage union
## ────────────────────────────────
## 1 1 2.000 11.000 0.160 0.000
## 2 1 3.000 -3.000 0.276 0.000
## 3 1 4.000 -1.000 0.000 0.000
## 4 1 5.000 3.000 0.065 0.000
## 5 1 6.000 -7.000 0.112 0.000
## 6 1 7.000 -3.000 0.070 0.000
## 7 2 2.000 -7.000 0.051 0.000
## 8 2 3.000 6.000 0.049 1.000
## 9 2 4.000 -3.000 0.281 -1.000
## 10 2 5.000 0.000 0.153 0.000
## 11 2 6.000 7.000 0.094 0.000
## 12 2 7.000 -7.000 0.024 0.000
## 13 3 2.000 1.000 0.784 0.000
## 14 3 3.000 -1.000 0.112 0.000
## 15 3 4.000 2.000 0.054 0.000
## 16 3 5.000 0.000 0.093 0.000
## 17 3 6.000 0.000 0.083 0.000
## 18 3 7.000 -6.000 0.082 0.000
## 19 4 2.000 -6.000 0.081 0.000
## 20 4 3.000 0.000 0.062 0.000
## ────────────────────────────────2.2 Wide and long format transfer
2.2.1 widen_panel()
## ───────────────────
## id t exp
## ───────────────────
## 1 1 1.000 3.000
## 2 1 2.000 4.000
## 3 1 3.000 5.000
## 4 1 4.000 6.000
## 5 1 5.000 7.000
## 6 1 6.000 8.000
## 7 1 7.000 9.000
## 8 2 1.000 30.000
## 9 2 2.000 31.000
## 10 2 3.000 32.000
## 11 2 4.000 33.000
## 12 2 5.000 34.000
## 13 2 6.000 35.000
## 14 2 7.000 36.000
## 15 3 1.000 6.000
## 16 3 2.000 7.000
## 17 3 3.000 8.000
## 18 3 4.000 9.000
## 19 3 5.000 10.000
## 20 3 6.000 11.000
## ───────────────────## ───────────────────────────────────────────────────────
## id exp_1 exp_2 exp_3 exp_4 exp_5 exp_6 exp_7
## ───────────────────────────────────────────────────────
## 1 1 3.000 4.000 5.000 6.000 7.000 8.000 9.000
## 2 2 30.000 31.000 32.000 33.000 34.000 35.000 36.000
## 3 3 6.000 7.000 8.000 9.000 10.000 11.000 12.000
## 4 4 31.000 32.000 33.000 34.000 35.000 36.000 37.000
## 5 5 10.000 11.000 12.000 13.000 14.000 15.000 16.000
## 6 6 26.000 27.000 28.000 29.000 30.000 31.000 32.000
## 7 7 15.000 16.000 17.000 18.000 19.000 20.000 21.000
## 8 8 23.000 24.000 25.000 26.000 27.000 28.000 29.000
## 9 9 3.000 4.000 5.000 6.000 7.000 8.000 9.000
## 10 10 3.000 4.000 5.000 6.000 7.000 8.000 9.000
## 11 11 24.000 25.000 26.000 27.000 28.000 29.000 30.000
## 12 12 21.000 22.000 23.000 24.000 25.000 26.000 27.000
## 13 13 26.000 27.000 28.000 29.000 30.000 31.000 32.000
## 14 14 15.000 16.000 17.000 18.000 19.000 20.000 21.000
## 15 15 9.000 10.000 11.000 12.000 13.000 14.000 15.000
## 16 16 16.000 17.000 18.000 19.000 20.000 21.000 22.000
## 17 17 16.000 17.000 18.000 19.000 20.000 21.000 22.000
## 18 18 25.000 26.000 27.000 28.000 29.000 30.000 31.000
## 19 19 40.000 41.000 42.000 43.000 44.000 45.000 46.000
## 20 20 25.000 26.000 27.000 28.000 29.000 30.000 31.000
## ───────────────────────────────────────────────────────2.2.2 long_panel()
long_wages <- long_panel(wide_wages, prefix = "_", begin = 1, end = 7,
id = "id", label_location = "end")
head(long_wages,20)%>%print_table()## ───────────────────
## id wave exp
## ───────────────────
## 1 1 1.000 3.000
## 2 1 2.000 4.000
## 3 1 3.000 5.000
## 4 1 4.000 6.000
## 5 1 5.000 7.000
## 6 1 6.000 8.000
## 7 1 7.000 9.000
## 8 2 1.000 30.000
## 9 2 2.000 31.000
## 10 2 3.000 32.000
## 11 2 4.000 33.000
## 12 2 5.000 34.000
## 13 2 6.000 35.000
## 14 2 7.000 36.000
## 15 3 1.000 6.000
## 16 3 2.000 7.000
## 17 3 3.000 8.000
## 18 3 4.000 9.000
## 19 3 5.000 10.000
## 20 3 6.000 11.000
## ───────────────────2.3 Checking data
2.3.1 summary(): Similar to Describe()
## ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## skim_type skim_variable t n_missing complete_rate numeric.mean numeric.sd numeric.p0 numeric.p25 numeric.p50 numeric.p75 numeric.p100 numeric.hist
## ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## 1 numeric lwage 1.000 0.000 1.000 6.375 0.388 5.011 6.116 6.425 6.653 6.906 ▁▂▃▇▇
## 2 numeric lwage 2.000 0.000 1.000 6.465 0.363 5.011 6.235 6.534 6.753 6.906 ▁▁▂▅▇
## 3 numeric lwage 3.000 0.000 1.000 6.597 0.447 4.605 6.333 6.613 6.856 8.269 ▁▂▇▃▁
## 4 numeric lwage 4.000 0.000 1.000 6.696 0.441 5.081 6.436 6.715 6.957 8.517 ▁▃▇▂▁
## 5 numeric lwage 5.000 0.000 1.000 6.786 0.424 5.273 6.515 6.805 7.038 8.102 ▁▂▇▅▁
## 6 numeric lwage 6.000 0.000 1.000 6.864 0.424 5.659 6.597 6.908 7.110 8.161 ▁▃▇▃▁
## 7 numeric lwage 7.000 0.000 1.000 6.951 0.438 5.677 6.685 6.985 7.208 8.537 ▁▅▇▂▁
## 8 numeric exp 1.000 0.000 1.000 16.854 10.790 1.000 7.000 15.000 26.000 45.000 ▇▇▅▅▁
## 9 numeric exp 2.000 0.000 1.000 17.854 10.790 2.000 8.000 16.000 27.000 46.000 ▇▇▅▅▁
## 10 numeric exp 3.000 0.000 1.000 18.854 10.790 3.000 9.000 17.000 28.000 47.000 ▇▇▅▅▁
## 11 numeric exp 4.000 0.000 1.000 19.854 10.790 4.000 10.000 18.000 29.000 48.000 ▇▇▅▅▁
## 12 numeric exp 5.000 0.000 1.000 20.854 10.790 5.000 11.000 19.000 30.000 49.000 ▇▇▅▅▁
## 13 numeric exp 6.000 0.000 1.000 21.854 10.790 6.000 12.000 20.000 31.000 50.000 ▇▇▅▅▁
## 14 numeric exp 7.000 0.000 1.000 22.854 10.790 7.000 13.000 21.000 32.000 51.000 ▇▇▅▅▁
## 15 numeric wks 1.000 0.000 1.000 46.281 6.253 6.000 46.000 48.000 50.000 52.000 ▁▁▁▁▇
## 16 numeric wks 2.000 0.000 1.000 47.020 5.135 11.000 47.000 49.000 50.000 52.000 ▁▁▁▁▇
## 17 numeric wks 3.000 0.000 1.000 47.045 4.767 20.000 47.000 49.000 50.000 52.000 ▁▁▁▁▇
## 18 numeric wks 4.000 0.000 1.000 47.192 4.463 8.000 47.000 48.000 50.000 52.000 ▁▁▁▁▇
## 19 numeric wks 5.000 0.000 1.000 46.961 4.890 6.000 47.000 48.000 50.000 52.000 ▁▁▁▁▇
## 20 numeric wks 6.000 0.000 1.000 46.729 4.982 6.000 46.000 48.000 50.000 52.000 ▁▁▁▁▇
## 21 numeric wks 7.000 0.000 1.000 46.452 5.185 5.000 46.000 48.000 49.000 52.000 ▁▁▁▁▇
## ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────2.3.2 are_varying(): Check if variables are constant or variable over time.
## occ ind fem blk
## TRUE TRUE FALSE FALSE
3 ANALYSIS
3.1 wbm(): CEM using MLM
使用由三部分组成的模型语法，如下所示：
dv ~ varying_variables | invariant_variables | cross_level_interactions/random effects
3.1.1 lag()
该函数还支持滞后变量lag()。与基本
R 不同，panelr正确滞后变量 - 第 1 波观测值将具有滞后变量的
NA 值，而不是采用前一个实体的最终波值。
model <- wbm(lwage ~ lag(union) + wks | blk | blk * wks + (lag(union) | id), data = wages)
summary(model)## MODEL INFO:
## Entities: 595
## Time periods: 2-7
## Dependent variable: lwage
## Model type: Linear mixed effects
## Specification: within-between
##
## MODEL FIT:
## AIC = 1427.04, BIC = 1495.03
## Pseudo-R² (fixed effects) = 0.05
## Pseudo-R² (total) = 0.75
## Entity ICC = 0.73
##
## WITHIN EFFECTS:
## ---------------------------------------------------------
## Est. S.E. t val. d.f. p
## ---------------- ------- ------ -------- --------- ------
## lag(union) 0.04 0.04 1.24 88.16 0.22
## wks -0.00 0.00 -1.51 2948.04 0.13
## ---------------------------------------------------------
##
## BETWEEN EFFECTS:
## ---------------------------------------------------------------
## Est. S.E. t val. d.f. p
## ----------------------- ------- ------ -------- -------- ------
## (Intercept) 6.20 0.24 25.89 571.97 0.00
## imean(lag(union)) 0.03 0.04 0.72 593.25 0.47
## imean(wks) 0.01 0.01 2.30 571.28 0.02
## blk -0.35 0.06 -5.65 591.86 0.00
## ---------------------------------------------------------------
##
## CROSS-LEVEL INTERACTIONS:
## ------------------------------------------------------
## Est. S.E. t val. d.f. p
## ------------- ------- ------ -------- --------- ------
## wks:blk -0.00 0.00 -1.06 2956.57 0.29
## ------------------------------------------------------
##
## p values calculated using Satterthwaite d.f.
##
## RANDOM EFFECTS:
## ------------------------------------
## Group Parameter Std. Dev.
## ---------- ------------- -----------
## id (Intercept) 0.3785
## id lag(union) 0.24
## Residual 0.2291
## ------------------------------------3.1.3 formula(): Get the formula from analysis
## lwage ~ lag(union) + wks | blk | blk * wks + (lag(union) | id)
## <environment: 0x000001a8cdfb48c0>3.1.4 Simple version of wbm()
## ────────────────────────────────────────────────────────────────────────
## group estimate std.error statistic p term
## ────────────────────────────────────────────────────────────────────────
## 1 within 0.053 0.025 2.132 .033 * lag(union)
## 2 within -0.001 0.001 -1.049 .294 wks
## 3 between 6.205 0.242 25.684 <.001 *** (Intercept)
## 4 between 0.025 0.036 0.687 .492 imean(lag(union))
## 5 between 0.011 0.005 2.267 .024 * imean(wks)
## 6 between -0.342 0.062 -5.525 <.001 *** blk
## 7 interactions -0.004 0.003 -1.212 .226 wks:blk
## 8 id 0.378 sd__(Intercept)
## 9 Residual 0.233 sd__Observation
## ────────────────────────────────────────────────────────────────────────## MODEL INFO:
## Entities: 595
## Time periods: 2-7
## Dependent variable: lwage
## Model type: Linear mixed effects
## Specification: within-between
##
## MODEL FIT:
## AIC = 1461.14, BIC = 1516.77
## Pseudo-R² (fixed effects) = 0.05
## Pseudo-R² (total) = 0.74
## Entity ICC = 0.73
##
## WITHIN EFFECTS:
## ---------------------------------------------------------
## Est. S.E. t val. d.f. p
## ---------------- ------- ------ -------- --------- ------
## lag(union) 0.05 0.03 2.13 2972.01 0.03
## wks -0.00 0.00 -1.05 2992.17 0.29
## ---------------------------------------------------------
##
## BETWEEN EFFECTS:
## ---------------------------------------------------------------
## Est. S.E. t val. d.f. p
## ----------------------- ------- ------ -------- -------- ------
## (Intercept) 6.21 0.24 25.68 590.99 0.00
## imean(lag(union)) 0.03 0.04 0.69 590.98 0.49
## imean(wks) 0.01 0.01 2.27 590.99 0.02
## blk -0.34 0.06 -5.53 591.03 0.00
## ---------------------------------------------------------------
##
## CROSS-LEVEL INTERACTIONS:
## ------------------------------------------------------
## Est. S.E. t val. d.f. p
## ------------- ------- ------ -------- --------- ------
## wks:blk -0.00 0.00 -1.21 2988.59 0.23
## ------------------------------------------------------
##
## p values calculated using Satterthwaite d.f.
##
## RANDOM EFFECTS:
## ------------------------------------
## Group Parameter Std. Dev.
## ---------- ------------- -----------
## id (Intercept) 0.3782
## Residual 0.2326
## ------------------------------------3.1.6 wbgee(): short version of wbm() with Linear GEE model
library(geepack)
wages <- panel_data(WageData, id = id, wave = t)
model <- wbgee(lwage ~ lag(union) + wks*blk, data = wages)
tidy(model)%>%print_table()## ────────────────────────────────────────────────────────────────────────
## group estimate std.error statistic p term
## ────────────────────────────────────────────────────────────────────────
## 1 within 0.014 0.020 0.737 .461 lag(union)
## 2 within -0.001 0.002 -0.634 .526 wks
## 3 between 6.231 0.249 24.989 <.001 *** (Intercept)
## 4 between 0.039 0.033 1.186 .236 imean(lag(union))
## 5 between 0.011 0.005 2.025 .043 * imean(wks)
## 6 between -0.334 0.061 -5.468 <.001 *** blk
## 7 interactions -0.001 0.002 -0.709 .478 wks:blk
## ────────────────────────────────────────────────────────────────────────## MODEL INFO:
## Entities: 595
## Time periods: 2-7
## Dependent variable: lwage
## Model type: Linear GEE
## Variance: ar1 (alpha = 0.86)
## Specification: within-between
##
## MODEL FIT:
## QIC = 717.9, QICu = 715.08, CIC = 8.41
##
## WITHIN EFFECTS:
## -----------------------------------------------
## Est. S.E. z val. p
## ---------------- ------- ------ -------- ------
## lag(union) 0.01 0.02 0.74 0.46
## wks -0.00 0.00 -0.63 0.53
## -----------------------------------------------
##
## BETWEEN EFFECTS:
## ------------------------------------------------------
## Est. S.E. z val. p
## ----------------------- ------- ------ -------- ------
## (Intercept) 6.23 0.25 24.99 0.00
## imean(lag(union)) 0.04 0.03 1.19 0.24
## imean(wks) 0.01 0.01 2.02 0.04
## blk -0.33 0.06 -5.47 0.00
## ------------------------------------------------------
##
## CROSS-LEVEL INTERACTIONS:
## --------------------------------------------
## Est. S.E. z val. p
## ------------- ------- ------ -------- ------
## wks:blk -0.00 0.00 -0.71 0.48
## --------------------------------------------3.1.7 fdm()
library(generics)
model <- fdm(lwage ~ lag(union) + wks*blk, data = wages)
generics::tidy(model)%>%print_table()## ─────────────────────────────────────────────────────
## estimate std.error statistic p term
## ─────────────────────────────────────────────────────
## 1 0.095 0.002 49.721 <.001 *** (Intercept)
## 2 0.023 0.017 1.320 .187 lag_union_
## 3 0.001 0.001 0.530 .597 wks
## 4 -0.002 0.002 -1.355 .176 wks:blk
## ─────────────────────────────────────────────────────## MODEL INFO:
## Entities: 595
## Time periods: 3-7
## Dependent variable: lwage
## Variance structure: toeplitz-1 (theta = -0.53)
##
## MODEL FIT:
## AIC = -1865.52, BIC = -1829.54
##
## Standard errors: CR2
## ------------------------------------------------
## Est. S.E. t val. p
## ----------------- ------- ------ -------- ------
## (Intercept) 0.10 0.00 49.72 0.00
## lag_union 0.02 0.02 1.32 0.19
## wks 0.00 0.00 0.53 0.60
## wks:blk -0.00 0.00 -1.36 0.18
## ------------------------------------------------3.2 Data generation
3.2.1 predict()
model <- wbm(lwage ~ lag(union) + wks*blk, data = wages)
# By default, assumes you're using the processed data for newdata
Pdata=predict(model)
head(Pdata,20)#%>%print_table()## 1 2 3 4 5 6 7 8 9 10 11 12 13
## 5.495 5.499 5.500 5.496 5.505 5.508 6.535 6.528 6.585 6.531 6.523 6.531 6.516
## 14 15 16 17 18 19 20
## 6.518 6.515 6.515 6.515 6.522 6.440 6.440