CRAN Package Check Results for Package fixest

Last updated on 2020-04-09 03:47:41 CEST.

Flavor Version Tinstall Tcheck Ttotal Status Flags
r-devel-linux-x86_64-debian-clang 0.4.0 94.94 102.38 197.32 OK
r-devel-linux-x86_64-debian-gcc 0.4.0 61.45 77.87 139.32 OK
r-devel-linux-x86_64-fedora-clang 0.4.0 260.55 NOTE
r-devel-linux-x86_64-fedora-gcc 0.4.0 236.31 OK
r-devel-windows-ix86+x86_64 0.4.0 132.00 197.00 329.00 OK
r-devel-windows-ix86+x86_64-gcc8 0.3.1 144.00 129.00 273.00 OK
r-patched-linux-x86_64 0.4.0 71.71 99.23 170.94 OK
r-patched-solaris-x86 0.4.0 264.60 OK
r-release-linux-x86_64 0.4.0 68.37 91.26 159.63 OK
r-release-windows-ix86+x86_64 0.4.0 151.00 176.00 327.00 OK
r-release-osx-x86_64 0.4.0 NOTE
r-oldrel-windows-ix86+x86_64 0.4.0 203.00 126.00 329.00 ERROR
r-oldrel-osx-x86_64 0.3.1 NOTE

Check Details

Version: 0.4.0
Check: installed package size
Result: NOTE
     installed size is 7.1Mb
     sub-directories of 1Mb or more:
     libs 4.8Mb
Flavors: r-devel-linux-x86_64-fedora-clang, r-release-osx-x86_64

Version: 0.4.0
Check: running examples for arch ‘i386’
Result: ERROR
    Running examples in 'fixest-Ex.R' failed
    The error most likely occurred in:
    
    > ### Name: collinearity
    > ### Title: Collinearity diagnostics for 'fixest' objects
    > ### Aliases: collinearity
    >
    > ### ** Examples
    >
    >
    > # Creating an example data base:
    > cluster_1 = sample(3, 100, TRUE)
    > cluster_2 = sample(20, 100, TRUE)
    > x = rnorm(100, cluster_1)**2
    > y = rnorm(100, cluster_2)**2
    > z = rnorm(100, 3)**2
    > dep = rpois(100, x*y*z)
    > base = data.frame(cluster_1, cluster_2, x, y, z, dep)
    >
    > # creating collinearity problems:
    > base$v1 = base$v2 = base$v3 = base$v4 = 0
    > base$v1[base$cluster_1 == 1] = 1
    > base$v2[base$cluster_1 == 2] = 1
    > base$v3[base$cluster_1 == 3] = 1
    > base$v4[base$cluster_2 == 1] = 1
    >
    > # Estimations:
    >
    > # Collinearity with the cluster variables:
    > res_1 = femlm(dep ~ log(x) + v1 + v2 + v4 | cluster_1 + cluster_2, base)
    Warning: [femlm]: The optimization algorithm did not converge, the results are not reliable. The information matrix is singular: presence of collinearity. Use function collinearity() to pinpoint the problems.
    > collinearity(res_1)
    [1] "Variables 'v1' and 'v2' are collinear with fixed-effects 'cluster_1'."
    > # => collinearity with cluster identified, we drop v1 and v2
    > res_1bis = femlm(dep ~ log(x) + v4 | cluster_1 + cluster_2, base)
    Warning: [femlm]: The information matrix is singular: presence of collinearity. Use function collinearity() to pinpoint the problems.
    > collinearity(res_1bis)
    [1] "Variable 'v4' is collinear with fixed-effects 'cluster_2'."
    >
    > # Multi-Collinearity:
    > res_2 = femlm(dep ~ log(x) + v1 + v2 + v3 + v4, base)
    Warning: [femlm]: The optimization algorithm did not converge, the results are not reliable. The information matrix is singular: presence of collinearity. Use function collinearity() to pinpoint the problems.
    > collinearity(res_2)
    [1] "Variable 'v1' is collinear with variables 'NA', 'v2' and 'v3'."
    >
    > # In non-linear part:
    > res_3 = feNmlm(dep ~ log(z), base, NL.fml = ~log(a*x + b*y),
    + NL.start = list(a=1, b=1), lower = list(a=0, b=0))
    Warning: [feNmlm] Optimization failed at iteration 13. Reason: Divergence... (evaluation of the log-likelihood is not finite)
    Try option verbose=2 to figure out the problem.
     Last evaluated coefficients returned.
    > collinearity(res_3)
    Error in collinearity(res_3) : Argument 'x' must be a fixest object.
    Execution halted
Flavor: r-oldrel-windows-ix86+x86_64

Version: 0.4.0
Check: running examples for arch ‘x64’
Result: ERROR
    Running examples in 'fixest-Ex.R' failed
    The error most likely occurred in:
    
    > ### Name: collinearity
    > ### Title: Collinearity diagnostics for 'fixest' objects
    > ### Aliases: collinearity
    >
    > ### ** Examples
    >
    >
    > # Creating an example data base:
    > cluster_1 = sample(3, 100, TRUE)
    > cluster_2 = sample(20, 100, TRUE)
    > x = rnorm(100, cluster_1)**2
    > y = rnorm(100, cluster_2)**2
    > z = rnorm(100, 3)**2
    > dep = rpois(100, x*y*z)
    > base = data.frame(cluster_1, cluster_2, x, y, z, dep)
    >
    > # creating collinearity problems:
    > base$v1 = base$v2 = base$v3 = base$v4 = 0
    > base$v1[base$cluster_1 == 1] = 1
    > base$v2[base$cluster_1 == 2] = 1
    > base$v3[base$cluster_1 == 3] = 1
    > base$v4[base$cluster_2 == 1] = 1
    >
    > # Estimations:
    >
    > # Collinearity with the cluster variables:
    > res_1 = femlm(dep ~ log(x) + v1 + v2 + v4 | cluster_1 + cluster_2, base)
    Warning: [femlm]: The optimization algorithm did not converge, the results are not reliable. The information matrix is singular: presence of collinearity. Use function collinearity() to pinpoint the problems.
    > collinearity(res_1)
    [1] "Variables 'v1' and 'v2' are collinear with fixed-effects 'cluster_1'."
    > # => collinearity with cluster identified, we drop v1 and v2
    > res_1bis = femlm(dep ~ log(x) + v4 | cluster_1 + cluster_2, base)
    Warning: [femlm]: The information matrix is singular: presence of collinearity. Use function collinearity() to pinpoint the problems.
    > collinearity(res_1bis)
    [1] "Variable 'v4' is collinear with fixed-effects 'cluster_2'."
    >
    > # Multi-Collinearity:
    > res_2 = femlm(dep ~ log(x) + v1 + v2 + v3 + v4, base)
    Warning: [femlm]: The information matrix is singular: presence of collinearity. Use function collinearity() to pinpoint the problems.
    > collinearity(res_2)
    [1] "Variable 'v1' is collinear with variables 'NA', 'v2' and 'v3'."
    >
    > # In non-linear part:
    > res_3 = feNmlm(dep ~ log(z), base, NL.fml = ~log(a*x + b*y),
    + NL.start = list(a=1, b=1), lower = list(a=0, b=0))
    Warning: [feNmlm] Optimization failed at iteration 13. Reason: Divergence... (evaluation of the log-likelihood is not finite)
    Try option verbose=2 to figure out the problem.
     Last evaluated coefficients returned.
    > collinearity(res_3)
    Error in collinearity(res_3) : Argument 'x' must be a fixest object.
    Execution halted
Flavor: r-oldrel-windows-ix86+x86_64

Version: 0.3.1
Check: installed package size
Result: NOTE
     installed size is 5.3Mb
     sub-directories of 1Mb or more:
     libs 3.1Mb
Flavor: r-oldrel-osx-x86_64