Columnar CSV file
Columnar CSV file
file name path
number of lines to skip, default=1
delimiter, default=','. Other options:
'\t': for tab-delimitd CSVRecordReader.QUOTE_HANDLING_DELIMITER: for records containing quotes with commas in them
quote, default='\"'
batch size, default=128
record reader name, default=rr
input columns, an array of tuples (indexFrom indexTo), default=Array()
one-hot input columns, an array of tuples (index, cardinality), default=Array()
output columns, an array of tuples (indexFrom indexTo), default=Array()
one-hot output columns, an array of tuples (index, cardinality), default=Array()
random generator seed, default=2018L
dataset iterator
Columnar CSV file
Columnar CSV file
file name path
columm name and types, preserving the order of the input files types: "double", "float", "long", "integer", "string", "categorical" (must provide the value list).
categorical column values
columns to ignore
columns to transform to one-hots
columns to transform to ints
number of lines to skip, default=1
delimiter, default=','. Other options:
'\t': for tab-delimitd CSVRecordReader.QUOTE_HANDLING_DELIMITER: for records containing quotes with commas in them
quote, default='\"'
batch size, default=128
record reader name, default=rr
input columns, an array of tuples (whether one-hot, (indexFrom indexTo)), default=Array()
output columns, an array of tuples (whether one-hot, (indexFrom indexTo)), default=Array()
dataset iterator
CSV Multi-Dataset Iterator