Open tillea opened 8 months ago
Just got the same issue on PowerPC 32-bit:
R version 4.3.3 (2024-02-29) -- "Angel Food Cake"
Copyright (C) 2024 The R Foundation for Statistical Computing
Platform: powerpc-apple-darwin10.0.0d2 (32-bit)
R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.
R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.
Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.
> library(testthat)
> library(estimatr)
>
> test_check("estimatr")
[ FAIL 4 | WARN 0 | SKIP 5 | PASS 1270 ]
══ Skipped tests (5) ═══════════════════════════════════════════════════════════
• On CRAN (3): 'test-iv-robust.R:235:3', 'test-lm-cluster.R:182:3',
'test-lm-cluster.R:213:3'
• ivpack cannot be loaded (2): 'test-iv-robust.R:30:3',
'test-iv-robust.R:328:3'
══ Failed tests ════════════════════════════════════════════════════════════════
── Failure ('test-lm-robust-fes.R:589:5'): FEs handle collinear covariates ─────
tidy(lmo)[grepl("cyl", lmo$term), ] not equivalent to tidy(lmfo)[grepl("cyl", lmfo$term), ].
Component "estimate": 'is.NA' value mismatch: 0 in current 1 in target
Component "std.error": 'is.NA' value mismatch: 0 in current 1 in target
Component "statistic": 'is.NA' value mismatch: 0 in current 1 in target
Component "p.value": 'is.NA' value mismatch: 0 in current 1 in target
Component "conf.low": 'is.NA' value mismatch: 0 in current 1 in target
Component "conf.high": 'is.NA' value mismatch: 0 in current 1 in target
Component "df": 'is.NA' value mismatch: 0 in current 1 in target
── Failure ('test-lm-robust-fes.R:589:5'): FEs handle collinear covariates ─────
tidy(lmo)[grepl("cyl", lmo$term), ] not equivalent to tidy(lmfo)[grepl("cyl", lmfo$term), ].
Component "estimate": 'is.NA' value mismatch: 0 in current 1 in target
Component "std.error": 'is.NA' value mismatch: 0 in current 1 in target
Component "statistic": 'is.NA' value mismatch: 0 in current 1 in target
Component "p.value": 'is.NA' value mismatch: 0 in current 1 in target
Component "conf.low": 'is.NA' value mismatch: 0 in current 1 in target
Component "conf.high": 'is.NA' value mismatch: 0 in current 1 in target
Component "df": 'is.NA' value mismatch: 0 in current 1 in target
── Failure ('test-lm-robust-fes.R:589:5'): FEs handle collinear covariates ─────
tidy(lmo)[grepl("cyl", lmo$term), ] not equivalent to tidy(lmfo)[grepl("cyl", lmfo$term), ].
Component "estimate": 'is.NA' value mismatch: 0 in current 1 in target
Component "std.error": 'is.NA' value mismatch: 0 in current 1 in target
Component "statistic": 'is.NA' value mismatch: 0 in current 1 in target
Component "p.value": 'is.NA' value mismatch: 0 in current 1 in target
Component "conf.low": 'is.NA' value mismatch: 0 in current 1 in target
Component "conf.high": 'is.NA' value mismatch: 0 in current 1 in target
Component "df": 'is.NA' value mismatch: 0 in current 1 in target
── Failure ('test-lm-robust-fes.R:622:3'): Handle perfect fits appropriately ───
tidy(rfo)[rfo$term == "X", ] not equivalent to tidy(ro)[ro$term == "X", ].
Component "std.error": 'is.NA' value mismatch: 1 in current 0 in target
Component "statistic": 'is.NA' value mismatch: 1 in current 0 in target
Component "p.value": 'is.NA' value mismatch: 1 in current 0 in target
Component "conf.low": 'is.NA' value mismatch: 1 in current 0 in target
Component "conf.high": 'is.NA' value mismatch: 1 in current 0 in target
[ FAIL 4 | WARN 0 | SKIP 5 | PASS 1270 ]
Error: Test failures
Execution halted
Any idea how to fix the test suite?
Any idea how to fix the test suite?
Probably expect_equal
should have tolerance being set. Something like in https://github.com/kevhuy/WALS/commit/5c8c57d825eb6fabc1d88b3d898f4ccf270e1557
P. S. Perhaps the issue is rather be renamed to 32-bit instead of i386.
Probably
expect_equal
should have tolerance being set. Something like in kevhuy/WALS@5c8c57d
Probably, but I did not wanted to do this without confirmation by upstream.
P. S. Perhaps the issue is rather be renamed to 32-bit instead of i386.
Not really since the testsuite is passing on armel and armhf. I've changed the issue title to what we really know.
Hi, when upgrading the Debian package of estimatr from version 1.0.0 to 1.0.2 a test suite error occurs on i386 architecture as you can see in the according test log. It contains
It would be great if you could have a look. Kind regards, Andreas.