Uploaded image for project: 'Apache Arrow'
  1. Apache Arrow
  2. ARROW-8899

[R] Add R metadata like pandas metadata for round-trip fidelity



    • Improvement
    • Status: Resolved
    • Critical
    • Resolution: Fixed
    • None
    • 1.0.0
    • R


      Arrow Schema and Field objects have custom_metadata fields to store arbitrary strings in a key-value store. Pandas stores JSON in a "pandas" key and uses that to improve the fidelity of round-tripping data to Arrow/Parquet/Feather and back. https://pandas.pydata.org/docs/dev/development/developer.html#storing-pandas-dataframe-objects-in-apache-parquet-format describes this a bit.

      You can see this pandas metadata in the sample Parquet file:

      tab <- read_parquet(system.file("v0.7.1.parquet", package="arrow"), as_data_frame = FALSE)
      # Table
      # 10 rows x 11 columns
      # $carat <double>
      # $cut <string>
      # $color <string>
      # $clarity <string>
      # $depth <double>
      # $table <double>
      # $price <int64>
      # $x <double>
      # $y <double>
      # $z <double>
      # $__index_level_0__ <int64>
      # $pandas
      # [1] "{\"index_columns\": [\"__index_level_0__\"], \"column_indexes\": [{\"name\": null, \"pandas_type\": \"string\", \"numpy_type\": \"object\", \"metadata\": null}], \"columns\": [{\"name\": \"carat\", \"pandas_type\": \"float64\", \"numpy_type\": \"float64\", \"metadata\": null}, {\"name\": \"cut\", \"pandas_type\": \"unicode\", \"numpy_type\": \"object\", \"metadata\": null}, {\"name\": \"color\", \"pandas_type\": \"unicode\", \"numpy_type\": \"object\", \"metadata\": null}, {\"name\": \"clarity\", \"pandas_type\": \"unicode\", \"numpy_type\": \"object\", \"metadata\": null}, {\"name\": \"depth\", \"pandas_type\": \"float64\", \"numpy_type\": \"float64\", \"metadata\": null}, {\"name\": \"table\", \"pandas_type\": \"float64\", \"numpy_type\": \"float64\", \"metadata\": null}, {\"name\": \"price\", \"pandas_type\": \"int64\", \"numpy_type\": \"int64\", \"metadata\": null}, {\"name\": \"x\", \"pandas_type\": \"float64\", \"numpy_type\": \"float64\", \"metadata\": null}, {\"name\": \"y\", \"pandas_type\": \"float64\", \"numpy_type\": \"float64\", \"metadata\": null}, {\"name\": \"z\", \"pandas_type\": \"float64\", \"numpy_type\": \"float64\", \"metadata\": null}, {\"name\": \"__index_level_0__\", \"pandas_type\": \"int64\", \"numpy_type\": \"int64\", \"metadata\": null}], \"pandas_version\": \"0.20.1\"}"

      We should do something similar in R: store the "attributes" for each column in a data.frame when we convert to Arrow, and restore those attributes when we read from Arrow.

      Since ARROW-8703, you could naively do this all in R, something like:

      tab$metadata$r <- lapply(df, attributes)

      on the conversion to Arrow, and in as.data.frame(), do

      if (!is.null(tab$metadata$r)) {
        df[] <- mapply(function(col, meta) {
          attributes(col) <- meta
        }, col = df, meta = tab$metadata$r)

      However, it's trickier than this because:

      • tab$metadata$r needs to be serialized to string and deserialized on the way back. Pandas uses JSON but arrow doesn't currently have a JSON R dependency. We could dput() to dump the R attributes, but that could introduce risks since you have to parse/eval code to consume it. My best idea at the moment is to try rawToChar(serialize(x, ascii = TRUE)) on the way out (ascii = TRUE doesn't mean it requires ASCII inputs, it's about how it serializes) and unserialize(charToRaw) on the way back. But maybe there's some lower-level way to do this better.
      • We'll need to do the same for all places where Tables and RecordBatches are created/converted
      • We'll need to make sure that nested types (structs) get the same coverage
      • This metadata only is attached to Schemas, meaning that Arrays/ChunkedArrays don't have a place to store extra metadata. So we probably want to attach to the R6 (Chunked)Array objects a metadata/attributes field so that if we convert an R vector to array, or if we extract an array out of a record batch, we don't lose the attributes.

      Doing this should resolve ARROW-4390 and make ARROW-8867 trivial as well.

      Finally, a note about this custom metadata vs. extension types. Extension types can be defined by adding metadata to a Field (in a Schema). I think this is out of scope here because we're only concerned with R roundtrip fidelity. If there were a type that (for example) R and Pandas both had that Arrow did not, we could define an extension type so that we could share that across the implementations. But unless/until there is value in establishing that extension type standard, let's not worry with it. (In other words, in R we should ignore pandas metadata; if there's anything that pandas wants to share with R, it will define it somewhere else.)


        Issue Links



              romainfrancois Romain Francois
              npr Neal Richardson
              0 Vote for this issue
              3 Start watching this issue



                Time Tracking

                  Original Estimate - Not Specified
                  Not Specified
                  Remaining Estimate - 0h
                  Time Spent - 2h 40m
                  2h 40m