Change Log

Next

• (spec-from) Attempt to create spec structure for sets of data

1.5.2

features

• Added CSV parsing for (transform-lines) that creates keys from first CSV row for map or defrecord structures on the remaining rows
• Added describe? to (shape) to indicate data structure types for keys
  Ex: {"a": {"b": {"c": 42}}} -> a{}.b{}.c [42]

bugs

• Fixed JSON key-flatten inconsistency in some nested structures

1.5.1

features

• Analyze the value types with regular expressions for each flattened path with counts. Ex: "a.b.c" -> {"ints" 34 "doubles" 21 "strings" 11}

1.5.0

features

• Flatten XML and/or JSON structures into dot delimited node names. Ex: {"a": {"b": {"c": 42}}} -> a.b.c [42]

1.4.1

• Changed name of (pre-process) to (disect) as a more appropriate name (not bw compatible, but within 24 hrs)
• more docs

1.4.0

• Added (pre-process) to pre-process input from larger data sets where only a subset of data is actually needed for analysis. Extract subsets of JSON with regex and before parsing.

1.3.3

• A collection optimization for (haystack) and (pivot-matrix*) with a significant performance increase depending on distribution of data/query diversity.