Unable to Generate Report

[janstadt]

Im using the latest R libs and RStudio and i've tried on 2 different macs and a windows box without any luck. On windows it seems to get the furthest, but ends up throwing this error:

label: unnamed-chunk-42
Quitting from lines 2045-2047 (Fantasy-Football-Analysis-v2.Rmd) 
Error in (function (cond)  : 
  error in evaluating the argument 'x' in selecting a method for function 'mean': subscript out of bounds
Calls: <Anonymous> ... eval -> mean -> as.numeric -> is.H2OFrame -> <Anonymous>

It does produce a csv file (dad_ranks.csv) as well as the graph images but fails to complete.

[sendmalwhere]

I see something similar intermittently. What kinds of inputs have you given for draftedOverall? I found that it works for me with no inputs, breaks with 1, and starts working again with more than 1.

[sendmalwhere]

My current theory is that the mean() function is making some bad assumptions in this index: [1:500,1]. But not being an R programmer, I'm a bit stumped by what to do to fix it.

[janstadt]

Hey @gatsbysghost. I havent changed any values from the current file. I see there are a bunch of random player names and teams in the calculations. Are we supposed to remove those or modify them as well?

[sendmalwhere]

That's a great question--part of me thinks they're essentially normalizers/validators for the data (e.g., "well, Jameis Winston should be about the mean value for points, so let's alter the table to reflect that assumption"), but on further examination I'm not so sure. I just removed that whole block with the mean functions and it did change the output CSV a little, but not much--and it also ran to completion. (The different CSV output might also be a result of my having re-run the whole model from top to bottom). So I might recommend just getting rid of them.

[sendmalwhere]

I'd have to defer to @jsoslow2 on their function--again, I'm not really sure they're mutating any data that gets used later, and it seems like they're just gumming up the works when there's a value out of range for the mean() function.

[daveashworth]

I'm actually getting an error when running on line 1620:

Error: Model-specific variable importance scores are currently not available for this type of model.
Execution halted

Running through RStudio

FYI I'm also an R newbie. Any thoughts are appreciated!

[sendmalwhere]

@daveashworth a recent commit fixes this issue (it's just a mis-named variable); pull from main again and you should be good to go!

[daveashworth]

Oh excellent thanks!