Clarify distinctions among table types

[lukejharmon]

Hi all! I am getting confused when making new functions and workflows by the different types of tables currently in the webapp (table:rows, table:r.dataframe, table:csv, and so on). It is not clear to me which one I should choose when, and why. Can we have a little write-up about this to help me sort things out? Also I believe there is a bug that I discovered, probably by doing the wrong thing. If you try to use table:rows for the input or output of an R function, you get a cryptic error (see attached). I am happy using dataframes only for R but this error was vague and it took me a while to sort it out.

Thanks!

[lukejharmon]

{
"inputs": [
{
"name": "table",
"type": "table",
"format": "rows"
},
{
"name": "ind_variable",
"type": "string",
"format": "text",
"domain": {
"format": "column.names",
"input": "table"
}
},
{
"name": "dep_variable",
"type": "string",
"format": "text",
"domain": {
"format": "column.names",
"input": "table"
}
}
],
"mode": "r",
"name": "minimum",
"outputs": [
{
"name": "tableResults",
"type": "table",
"format": "r.dataframe"
}
],
"script": "# get x and y data with names\n\nx <- table[,ind_variable]\nnames(x) <- rownames(table)\ny <- table[,dep_variable]\nnames(y) <- rownames(table)\n\n# bind x and y\ntableResults <- as.data.frame(cbind(x, y))\n",
"description": ""
}

[lukejharmon]

that is a minimal example. I think the problem is restricted to using table:rows on a "csv" object on the INPUT side. I get a more meaningful error if I mess with the outputs.