df <- read.csv('C:/Users/Asus/OneDrive/Рабочий стол/Итог_2.csv', sep = ';')
df$Cluster <- as.factor(df$Cluster)
tab_corr(df[ , c(10, 11, 12, 13, 14, 15 )], corr.method = "spearman")
|
|
Rang
|
ln_cap
|
ln_rev
|
ln_lab
|
ln_vrp
|
ln_inv
|
|
Rang
|
|
-0.241***
|
-0.373***
|
-0.335***
|
0.051
|
-0.439***
|
|
ln_cap
|
-0.241***
|
|
0.810***
|
0.744***
|
0.050
|
0.135***
|
|
ln_rev
|
-0.373***
|
0.810***
|
|
0.855***
|
0.044
|
0.200***
|
|
ln_lab
|
-0.335***
|
0.744***
|
0.855***
|
|
-0.028
|
0.130***
|
|
ln_vrp
|
0.051
|
0.050
|
0.044
|
-0.028
|
|
0.664***
|
|
ln_inv
|
-0.439***
|
0.135***
|
0.200***
|
0.130***
|
0.664***
|
|
|
Computed correlation used spearman-method with listwise-deletion.
|
corr_mat <- round(cor(df[ , c(10, 11, 12, 13, 14, 15 )], method = "spearman"), 2)
corrplot(corr_mat, method = 'number', tl.col = "black",
cl.ratio = 0.2, tl.srt = 45, col = COL2('PuOr', 10))
M1 <- plm(ln_rev ~ ln_lab + ln_cap + Rang + ln_inv + ln_vrp, df, model = "pooling")
## Warning in pdata.frame(data, index = index, ...): duplicate couples (id-time) in resulting pdata.frame
## to find out which, use, e.g., table(index(your_pdataframe), useNA = "ifany")
tab_model(M1)
|
|
ln rev
|
|
Predictors
|
Estimates
|
CI
|
p
|
|
(Intercept)
|
5.28
|
1.83 – 8.74
|
0.003
|
|
ln lab
|
0.76
|
0.72 – 0.81
|
<0.001
|
|
ln cap
|
0.27
|
0.24 – 0.30
|
<0.001
|
|
Rang
|
-0.01
|
-0.02 – -0.01
|
<0.001
|
|
ln inv
|
-0.29
|
-0.80 – 0.21
|
0.255
|
|
ln vrp
|
0.76
|
0.27 – 1.25
|
0.003
|
|
Observations
|
1392
|
|
R2 / R2 adjusted
|
0.794 / 0.793
|
summary(M1)
## Pooling Model
##
## Call:
## plm(formula = ln_rev ~ ln_lab + ln_cap + Rang + ln_inv + ln_vrp,
## data = df, model = "pooling")
##
## Unbalanced Panel: n = 348, T = 4-4, N = 1392
##
## Residuals:
## Min. 1st Qu. Median 3rd Qu. Max.
## -6.014992 -0.720341 -0.086666 0.633278 5.230226
##
## Coefficients:
## Estimate Std. Error t-value Pr(>|t|)
## (Intercept) 5.2848000 1.7601279 3.0025 0.002726 **
## ln_lab 0.7620624 0.0233637 32.6174 < 2.2e-16 ***
## ln_cap 0.2698373 0.0134507 20.0613 < 2.2e-16 ***
## Rang -0.0128339 0.0021419 -5.9919 2.641e-09 ***
## ln_inv -0.2933609 0.2577948 -1.1380 0.255333
## ln_vrp 0.7596887 0.2510426 3.0261 0.002523 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Total Sum of Squares: 9436.5
## Residual Sum of Squares: 1942.6
## R-Squared: 0.79413
## Adj. R-Squared: 0.79339
## F-statistic: 1069.31 on 5 and 1386 DF, p-value: < 2.22e-16
M2 <- plm(ln_rev ~ ln_lab+ ln_cap + Rang + ln_inv + ln_vrp, df, model = "pooling")
## Warning in pdata.frame(data, index = index, ...): duplicate couples (id-time) in resulting pdata.frame
## to find out which, use, e.g., table(index(your_pdataframe), useNA = "ifany")
tab_model(M2)
|
|
ln rev
|
|
Predictors
|
Estimates
|
CI
|
p
|
|
(Intercept)
|
5.28
|
1.83 – 8.74
|
0.003
|
|
ln lab
|
0.76
|
0.72 – 0.81
|
<0.001
|
|
ln cap
|
0.27
|
0.24 – 0.30
|
<0.001
|
|
Rang
|
-0.01
|
-0.02 – -0.01
|
<0.001
|
|
ln inv
|
-0.29
|
-0.80 – 0.21
|
0.255
|
|
ln vrp
|
0.76
|
0.27 – 1.25
|
0.003
|
|
Observations
|
1392
|
|
R2 / R2 adjusted
|
0.794 / 0.793
|
summary(M2)
## Pooling Model
##
## Call:
## plm(formula = ln_rev ~ ln_lab + ln_cap + Rang + ln_inv + ln_vrp,
## data = df, model = "pooling")
##
## Unbalanced Panel: n = 348, T = 4-4, N = 1392
##
## Residuals:
## Min. 1st Qu. Median 3rd Qu. Max.
## -6.014992 -0.720341 -0.086666 0.633278 5.230226
##
## Coefficients:
## Estimate Std. Error t-value Pr(>|t|)
## (Intercept) 5.2848000 1.7601279 3.0025 0.002726 **
## ln_lab 0.7620624 0.0233637 32.6174 < 2.2e-16 ***
## ln_cap 0.2698373 0.0134507 20.0613 < 2.2e-16 ***
## Rang -0.0128339 0.0021419 -5.9919 2.641e-09 ***
## ln_inv -0.2933609 0.2577948 -1.1380 0.255333
## ln_vrp 0.7596887 0.2510426 3.0261 0.002523 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Total Sum of Squares: 9436.5
## Residual Sum of Squares: 1942.6
## R-Squared: 0.79413
## Adj. R-Squared: 0.79339
## F-statistic: 1069.31 on 5 and 1386 DF, p-value: < 2.22e-16
tab_model(lm(ln_rev ~ (ln_lab + ln_cap + Rang + ln_inv + ln_vrp)*Cluster, df))
|
|
ln rev
|
|
Predictors
|
Estimates
|
CI
|
p
|
|
(Intercept)
|
6.04
|
0.11 – 11.97
|
0.046
|
|
ln lab
|
0.86
|
0.78 – 0.93
|
<0.001
|
|
ln cap
|
0.30
|
0.25 – 0.34
|
<0.001
|
|
Rang
|
-0.01
|
-0.02 – -0.00
|
0.024
|
|
ln inv
|
0.29
|
-0.49 – 1.07
|
0.469
|
|
ln vrp
|
0.12
|
-0.54 – 0.79
|
0.718
|
|
Cluster [1]
|
-10.99
|
-18.41 – -3.56
|
0.004
|
|
ln lab × Cluster [1]
|
-0.22
|
-0.31 – -0.13
|
<0.001
|
|
ln cap × Cluster [1]
|
-0.06
|
-0.11 – -0.01
|
0.032
|
|
Rang × Cluster [1]
|
0.02
|
0.01 – 0.04
|
<0.001
|
|
ln inv × Cluster [1]
|
-0.35
|
-1.38 – 0.68
|
0.508
|
|
ln vrp × Cluster [1]
|
1.25
|
0.31 – 2.19
|
0.009
|
|
Observations
|
1392
|
|
R2 / R2 adjusted
|
0.813 / 0.812
|
tab_model(M1, M2)
|
|
ln rev
|
ln rev
|
|
Predictors
|
Estimates
|
CI
|
p
|
Estimates
|
CI
|
p
|
|
(Intercept)
|
5.28
|
1.83 – 8.74
|
0.003
|
5.28
|
1.83 – 8.74
|
0.003
|
|
ln lab
|
0.76
|
0.72 – 0.81
|
<0.001
|
0.76
|
0.72 – 0.81
|
<0.001
|
|
ln cap
|
0.27
|
0.24 – 0.30
|
<0.001
|
0.27
|
0.24 – 0.30
|
<0.001
|
|
Rang
|
-0.01
|
-0.02 – -0.01
|
<0.001
|
-0.01
|
-0.02 – -0.01
|
<0.001
|
|
ln inv
|
-0.29
|
-0.80 – 0.21
|
0.255
|
-0.29
|
-0.80 – 0.21
|
0.255
|
|
ln vrp
|
0.76
|
0.27 – 1.25
|
0.003
|
0.76
|
0.27 – 1.25
|
0.003
|
|
Observations
|
1392
|
1392
|
|
R2 / R2 adjusted
|
0.794 / 0.793
|
0.794 / 0.793
|
b <- plm(ln_rev ~ ln_lab + ln_cap + Rang + ln_inv + ln_vrp + Cluster, df, model = "random")
## Warning in pdata.frame(data, index = index, ...): duplicate couples (id-time) in resulting pdata.frame
## to find out which, use, e.g., table(index(your_pdataframe), useNA = "ifany")
summary(b)
## Oneway (individual) effect Random Effect Model
## (Swamy-Arora's transformation)
##
## Call:
## plm(formula = ln_rev ~ ln_lab + ln_cap + Rang + ln_inv + ln_vrp +
## Cluster, data = df, model = "random")
##
## Unbalanced Panel: n = 348, T = 4-4, N = 1392
##
## Effects:
## var std.dev share
## idiosyncratic 0.2663 0.5160 0.197
## individual 1.0847 1.0415 0.803
## theta:
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.7595 0.7595 0.7595 0.7595 0.7595 0.7595
##
## Residuals:
## Min. 1st Qu. Median 3rd Qu. Max.
## -4.5285694 -0.2305911 0.0033076 0.2672087 2.8258057
##
## Coefficients:
## Estimate Std. Error z-value Pr(>|z|)
## (Intercept) 4.2353247 1.4485712 2.9238 0.003458 **
## ln_lab 0.6483853 0.0297078 21.8255 < 2.2e-16 ***
## ln_cap 0.1966042 0.0131588 14.9409 < 2.2e-16 ***
## Rang 0.0033984 0.0034570 0.9830 0.325587
## ln_inv -0.0339248 0.1646271 -0.2061 0.836736
## ln_vrp 0.6440594 0.1448221 4.4472 8.698e-06 ***
## Cluster1 1.0522486 0.1713881 6.1396 8.275e-10 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Total Sum of Squares: 885.92
## Residual Sum of Squares: 395.54
## R-Squared: 0.55353
## Adj. R-Squared: 0.5516
## Chisq: 1717.11 on 6 DF, p-value: < 2.22e-16
a <- plm(ln_rev ~ ln_lab + ln_cap + Rang + ln_inv + ln_vrp + Cluster, df, model = "within")
## Warning in pdata.frame(data, index = index, ...): duplicate couples (id-time) in resulting pdata.frame
## to find out which, use, e.g., table(index(your_pdataframe), useNA = "ifany")
summary(a)
## Oneway (individual) effect Within Model
##
## Call:
## plm(formula = ln_rev ~ ln_lab + ln_cap + Rang + ln_inv + ln_vrp +
## Cluster, data = df, model = "within")
##
## Unbalanced Panel: n = 348, T = 4-4, N = 1392
##
## Residuals:
## Min. 1st Qu. Median 3rd Qu. Max.
## -4.3405697 -0.1466563 0.0053268 0.1611811 2.7718635
##
## Coefficients:
## Estimate Std. Error t-value Pr(>|t|)
## ln_lab 0.3697854 0.0432631 8.5474 < 2.2e-16 ***
## ln_cap 0.1490598 0.0146512 10.1739 < 2.2e-16 ***
## Rang 0.0121956 0.0057544 2.1194 0.0343 *
## ln_inv -0.1002592 0.1854226 -0.5407 0.5888
## ln_vrp 0.8386919 0.1682776 4.9840 7.294e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Total Sum of Squares: 361.13
## Residual Sum of Squares: 276.67
## R-Squared: 0.23389
## Adj. R-Squared: -0.02566
## F-statistic: 63.4399 on 5 and 1039 DF, p-value: < 2.22e-16
feglm(ln_rev ~ ln_lab+ ln_cap + Rang + ln_inv + ln_vrp + Cluster, df, fixef_vars = "Cluster")
## Warning: In fixest_env(fml = fml, data = data, family = famil...:
## feglm(fml = ln_...: fixef_vars is not a valid argument for function feglm()
## GLM estimation, family = gaussian, Dep. Var.: ln_rev
## Observations: 1,392
## Standard-errors: IID
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.869444 1.760484 2.197944 2.8118e-02 *
## ln_lab 0.737412 0.023530 31.339211 < 2.2e-16 ***
## ln_cap 0.263726 0.013365 19.732580 < 2.2e-16 ***
## Rang -0.002239 0.002800 -0.799520 4.2413e-01
## ln_inv -0.131778 0.256872 -0.513010 6.0803e-01
## ln_vrp 0.686761 0.248984 2.758253 5.8876e-03 **
## Cluster1 0.591398 0.102036 5.795982 8.3988e-09 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## Log-Likelihood: -2,191.4 Adj. Pseudo R2: 0.335773
## BIC: 4,433.5 Squared Cor.: 0.79903
tab_model(feglm(ln_rev ~ ln_lab+ ln_cap + Rang + ln_inv + ln_vrp + relevel(Cluster, ref = "1"), df))
|
|
ln rev
|
|
Predictors
|
Estimates
|
CI
|
p
|
|
(Intercept)
|
4.46
|
1.03 – 7.89
|
0.011
|
|
ln lab
|
0.74
|
0.69 – 0.78
|
<0.001
|
|
ln cap
|
0.26
|
0.24 – 0.29
|
<0.001
|
|
Rang
|
-0.00
|
-0.01 – 0.00
|
0.424
|
|
ln inv
|
-0.13
|
-0.64 – 0.37
|
0.608
|
|
ln vrp
|
0.69
|
0.20 – 1.18
|
0.006
|
relevel(Cluster, ref =
“1”)0
|
-0.59
|
-0.79 – -0.39
|
<0.001
|
|
Observations
|
1392
|
|
R2 / R2 adjusted
|
0.338 / 0.336
|
tab_model(feglm(ln_rev ~ ln_lab+ ln_cap + Rang + ln_inv + ln_vrp + Cluster, df))
|
|
ln rev
|
|
Predictors
|
Estimates
|
CI
|
p
|
|
(Intercept)
|
3.87
|
0.42 – 7.32
|
0.028
|
|
ln lab
|
0.74
|
0.69 – 0.78
|
<0.001
|
|
ln cap
|
0.26
|
0.24 – 0.29
|
<0.001
|
|
Rang
|
-0.00
|
-0.01 – 0.00
|
0.424
|
|
ln inv
|
-0.13
|
-0.64 – 0.37
|
0.608
|
|
ln vrp
|
0.69
|
0.20 – 1.18
|
0.006
|
|
Cluster [1]
|
0.59
|
0.39 – 0.79
|
<0.001
|
|
Observations
|
1392
|
|
R2 / R2 adjusted
|
0.338 / 0.336
|
phtest(a, b)
##
## Hausman Test
##
## data: ln_rev ~ ln_lab + ln_cap + Rang + ln_inv + ln_vrp + Cluster
## chisq = 111.76, df = 5, p-value < 2.2e-16
## alternative hypothesis: one model is inconsistent
zz <- plm(log(Revenue) ~ log(Capital) + log(Labor) + log(VRP) + log(Invest) + Cluster, data = df)
## Warning in pdata.frame(data, index = index, ...): duplicate couples (id-time) in resulting pdata.frame
## to find out which, use, e.g., table(index(your_pdataframe), useNA = "ifany")
zzwith <- zz$within
zzra <- plm(log(Revenue) ~ log(Capital) + log(Labor) + log(VRP) + log(Invest) + Cluster, data = df, model = "random")
## Warning in pdata.frame(data, index = index, ...): duplicate couples (id-time) in resulting pdata.frame
## to find out which, use, e.g., table(index(your_pdataframe), useNA = "ifany")
print(zzra)
##
## Model Formula: log(Revenue) ~ log(Capital) + log(Labor) + log(VRP) + log(Invest) +
## Cluster
##
## Coefficients:
## (Intercept) log(Capital) log(Labor) log(VRP) log(Invest) Cluster1
## 4.815862 0.197752 0.648535 0.615764 -0.037874 0.941960
print(summary(zzra))
## Oneway (individual) effect Random Effect Model
## (Swamy-Arora's transformation)
##
## Call:
## plm(formula = log(Revenue) ~ log(Capital) + log(Labor) + log(VRP) +
## log(Invest) + Cluster, data = df, model = "random")
##
## Unbalanced Panel: n = 348, T = 3-4, N = 1391
##
## Effects:
## var std.dev share
## idiosyncratic 0.2672 0.5169 0.198
## individual 1.0830 1.0407 0.802
## theta:
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.7244 0.7590 0.7590 0.7589 0.7590 0.7590
##
## Residuals:
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -4.5218 -0.2265 0.0027 0.0000 0.2689 2.8330
##
## Coefficients:
## Estimate Std. Error z-value Pr(>|z|)
## (Intercept) 4.815862 1.351734 3.5627 0.000367 ***
## log(Capital) 0.197752 0.013118 15.0745 < 2.2e-16 ***
## log(Labor) 0.648535 0.029686 21.8462 < 2.2e-16 ***
## log(VRP) 0.615764 0.141518 4.3511 1.354e-05 ***
## log(Invest) -0.037874 0.164787 -0.2298 0.818217
## Cluster1 0.941960 0.130309 7.2287 4.877e-13 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Total Sum of Squares: 888.24
## Residual Sum of Squares: 395.95
## R-Squared: 0.55423
## Adj. R-Squared: 0.55262
## Chisq: 1720.64 on 5 DF, p-value: < 2.22e-16
print(summary(zz))
## Oneway (individual) effect Within Model
##
## Call:
## plm(formula = log(Revenue) ~ log(Capital) + log(Labor) + log(VRP) +
## log(Invest) + Cluster, data = df)
##
## Unbalanced Panel: n = 348, T = 3-4, N = 1391
##
## Residuals:
## Min. 1st Qu. Median 3rd Qu. Max.
## -4.3666458 -0.1428966 0.0065714 0.1614145 2.7517449
##
## Coefficients:
## Estimate Std. Error t-value Pr(>|t|)
## log(Capital) 0.150698 0.014655 10.2828 < 2.2e-16 ***
## log(Labor) 0.370587 0.043332 8.5522 < 2.2e-16 ***
## log(VRP) 0.646678 0.141300 4.5766 5.295e-06 ***
## log(Invest) 0.037644 0.172403 0.2183 0.8272
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Total Sum of Squares: 360.62
## Residual Sum of Squares: 277.58
## R-Squared: 0.23029
## Adj. R-Squared: -0.029734
## F-statistic: 77.7159 on 4 and 1039 DF, p-value: < 2.22e-16