R: Pander sink stack full when printing summary lm

问题

I am in the middle of generating a HTML report in Rstudio via pandoc for a collaborator. However pander is hitting the sink limit in R when trying to generate the output for the following summary of a lm() object.

My R instance:

version _
platform x86_64-apple-darwin13.1.0
arch x86_64
os darwin13.1.0
system x86_64, darwin13.1.0
status
major 3
minor 1.0
year 2014
month 04
day 10
svn rev 65387
language R
version.string R version 3.1.0 (2014-04-10) nickname Spring Dance

The lm I am trying to generate the HTML from:

summary(a.lm)

Call:
lm(formula = f.t ~ ((b.c + I(a.c^2) + b.t.c + 
    c.c) * (PTGENDER + DX.bl + Al1.Al2)) + BMI.t.c + FH.t.c + 
    Hb.t.c + age.c + I(age.c^2), data = subset(d, 
    !is.na(Hb.t.c) & !is.na(BMI.t.c) & !is.na(b.t.c) & !is.na(c.c)))

Residuals:
     Min       1Q   Median       3Q      Max 
-0.72042 -0.15865 -0.01311  0.15826  0.91421 

Coefficients:
                             Estimate Std. Error t value Pr(>|t|)    
(Intercept)                 1.6247957  0.1120717  14.498  < 2e-16 ***
a.c                         0.1343279  0.1481459   0.907  0.36541    
I(a.c^2)                   -0.0312254  0.0537683  -0.581  0.56193    
b.t.c                       0.1040226  0.0782932   1.329  0.18516    
c.c                         0.2012898  0.0818699   2.459  0.01461 *  
PTGENDERFemale              0.0067323  0.0473586   0.142  0.88707    
DX.blLMCI                  -0.0121387  0.0553348  -0.219  0.82654    
DX.blAD                     0.0209534  0.0769331   0.272  0.78557    
Al1.Al23:3                  0.1658594  0.1075438   1.542  0.12426    
Al1.Al23:4                  0.3506150  0.1121648   3.126  0.00198 ** 
Al1.Al24:4                  0.1386058  0.1690946   0.820  0.41316    
BMI.t.c                    -0.0280475  0.0182582  -1.536  0.12574    
FH.t.c                      0.0031254  0.0225851   0.138  0.89005    
Hb.t.c                      0.0162085  0.0171502   0.945  0.34551    
age.c                       0.0024730  0.0028751   0.860  0.39052    
I(age.c^2)                  0.0003961  0.0002926   1.354  0.17706    
a.c:PTGENDERFemale          0.0303573  0.0474341   0.640  0.52276    
a.c:DX.blLMCI               0.1143316  0.0603606   1.894  0.05934 .  
a.c:DX.blAD                 0.1406856  0.0691761   2.034  0.04302 *  
a.c:Al1.Al23:3             -0.0369030  0.1472692  -0.251  0.80234    
a.c:Al1.Al23:4             -0.0493593  0.1469924  -0.336  0.73730    
a.c:Al1.Al24:4             -0.0688362  0.1564964  -0.440  0.66041    
I(a.c^2):PTGENDERFemale    -0.0584777  0.0324483  -1.802  0.07270 .  
I(a.c^2):DX.blLMCI          0.0298364  0.0368905   0.809  0.41940    
I(a.c^2):DX.blAD            0.0147823  0.0440503   0.336  0.73747    
I(a.c^2):Al1.Al23:3         0.0001684  0.0586422   0.003  0.99771    
I(a.c^2):Al1.Al23:4        -0.0357507  0.0621727  -0.575  0.56578    
I(a.c^2):Al1.Al24:4        -0.0328713  0.0664954  -0.494  0.62149    
b.t.c:PTGENDERFemale        0.0208930  0.0466606   0.448  0.65470    
b.t.c:DX.blLMCI            -0.0640037  0.0628767  -1.018  0.30968    
b.t.c:DX.blAD              -0.0607185  0.0705763  -0.860  0.39042    
b.t.c:Al1.Al23:3           -0.0428761  0.0727674  -0.589  0.55624    
b.t.c:Al1.Al23:4           -0.0289522  0.0729465  -0.397  0.69178    
b.t.c:Al1.Al24:4           -0.0946543  0.0981224  -0.965  0.33564    
c.c:PTGENDERFemale          0.0193275  0.0445454   0.434  0.66474    
c.c:DX.blLMCI              -0.0371738  0.0470631  -0.790  0.43034    
c.c:DX.blAD                -0.1221799  0.0722145  -1.692  0.09189 .  
c.c:Al1.Al23:3             -0.1604799  0.0798383  -2.010  0.04548 *  
c.c:Al1.Al23:4             -0.2015401  0.0871098  -2.314  0.02149 *  
c.c:Al1.Al24:4             -0.3374309  0.1520609  -2.219  0.02737 *  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.2795 on 254 degrees of freedom
Multiple R-squared:  0.4536,    Adjusted R-squared:  0.3697 
F-statistic: 5.406 on 39 and 254 DF,  p-value: < 2.2e-16

running:

pander(summary(a.lm))

leads to the following error

Error in sink(file) : sink stack is full

Any suggestions to overcome this would be much appreciated.

Cheers

In response to Mr Flick

Yes it seems to:

Here is an example:

d <- data.frame(a = rnorm(20), b = sample(letters[1:2], 20, replace = TRUE))
pander(summary(lm(a ~ b, data = d)))
Error in sink(file) : sink stack is full
pander(lm(a ~ b, data = d))
--------------------------------------------------------------
     &nbsp;        Estimate   Std. Error   t value   Pr(>|t|) 
----------------- ---------- ------------ --------- ----------
     **bb**        -0.1637      0.3797     -0.4311    0.6715  

 **(Intercept)**   -0.1634      0.2401     -0.6805    0.5049  
--------------------------------------------------------------

Table: Fitting linear model: a ~ b

So dropping the summary call works :)

回答1:

Actually, if you're just running a standard lm regression, don't pass the summary.lm object to pander, pass the model itself (the lm object)

pander(a.lm)

That should have the information you need. The function calls summary() internally.

来源：https://stackoverflow.com/questions/24359744/r-pander-sink-stack-full-when-printing-summary-lm

标签

sink

pander