Module: Polars::IO

Included in:
Polars
Defined in:
lib/polars/io/csv.rb,
lib/polars/io/ipc.rb,
lib/polars/io/avro.rb,
lib/polars/io/json.rb,
lib/polars/io/ndjson.rb,
lib/polars/io/parquet.rb,
lib/polars/io/database.rb

Instance Method Summary collapse

Instance Method Details

#read_avro(source, columns: nil, n_rows: nil) ⇒ DataFrame

Read into a DataFrame from Apache Avro format.

Parameters:

  • source (Object) —

    Path to a file or a file-like object.

  • columns (Object) (defaults to: nil) —

    Columns to select. Accepts a list of column indices (starting at zero) or a list of column names.

  • n_rows (Integer) (defaults to: nil) —

    Stop reading from Apache Avro file after reading n_rows.

Returns:



14
15
16
17
18
19
20
21
22
 
# File 'lib/polars/io/avro.rb', line 14

def read_avro(source, columns: nil, n_rows: nil)
  if Utils.pathlike?(source)
    source = Utils.normalize_filepath(source)
  end
  projection, column_names = Utils.handle_projection_columns(columns)

  rbdf = RbDataFrame.read_avro(source, column_names, projection, n_rows)
  Utils.wrap_df(rbdf)
end

#read_csv(source, has_header: true, columns: nil, new_columns: nil, sep: ",", comment_char: nil, quote_char: '"', skip_rows: 0, dtypes: nil, null_values: nil, ignore_errors: false, parse_dates: false, n_threads: nil, infer_schema_length: N_INFER_DEFAULT, batch_size: 8192, n_rows: nil, encoding: "utf8", low_memory: false, rechunk: true, storage_options: nil, skip_rows_after_header: 0, row_count_name: nil, row_count_offset: 0, eol_char: "\n", truncate_ragged_lines: false) ⇒ DataFrame

Note:

This operation defaults to a rechunk operation at the end, meaning that all data will be stored continuously in memory. Set rechunk: false if you are benchmarking the csv-reader. A rechunk is an expensive operation.

Read a CSV file into a DataFrame.

Parameters:

  • source (Object) —

    Path to a file or a file-like object.

  • has_header (Boolean) (defaults to: true) —

    Indicate if the first row of dataset is a header or not. If set to false, column names will be autogenerated in the following format: column_x, with x being an enumeration over every column in the dataset starting at 1.

  • columns (Object) (defaults to: nil) —

    Columns to select. Accepts a list of column indices (starting at zero) or a list of column names.

  • new_columns (Object) (defaults to: nil) —

    Rename columns right after parsing the CSV file. If the given list is shorter than the width of the DataFrame the remaining columns will have their original name.

  • sep (String) (defaults to: ",") —

    Single byte character to use as delimiter in the file.

  • comment_char (String) (defaults to: nil) —

    Single byte character that indicates the start of a comment line, for instance #.

  • quote_char (String) (defaults to: '"') —

    Single byte character used for csv quoting. Set to nil to turn off special handling and escaping of quotes.

  • skip_rows (Integer) (defaults to: 0) —

    Start reading after skip_rows lines.

  • dtypes (Object) (defaults to: nil) —

    Overwrite dtypes during inference.

  • null_values (Object) (defaults to: nil) —

    Values to interpret as null values. You can provide a:

    • String: All values equal to this string will be null.
    • Array: All values equal to any string in this array will be null.
    • Hash: A hash that maps column name to a null value string.
  • ignore_errors (Boolean) (defaults to: false) —

    Try to keep reading lines if some lines yield errors. First try infer_schema_length: 0 to read all columns as :str to check which values might cause an issue.

  • parse_dates (Boolean) (defaults to: false) —

    Try to automatically parse dates. If this does not succeed, the column remains of data type :str.

  • n_threads (Integer) (defaults to: nil) —

    Number of threads to use in csv parsing. Defaults to the number of physical cpu's of your system.

  • infer_schema_length (Integer) (defaults to: N_INFER_DEFAULT) —

    Maximum number of lines to read to infer schema. If set to 0, all columns will be read as :utf8. If set to nil, a full table scan will be done (slow).

  • batch_size (Integer) (defaults to: 8192) —

    Number of lines to read into the buffer at once. Modify this to change performance.

  • n_rows (Integer) (defaults to: nil) —

    Stop reading from CSV file after reading n_rows. During multi-threaded parsing, an upper bound of n_rows rows cannot be guaranteed.

  • encoding ("utf8", "utf8-lossy") (defaults to: "utf8") —

    Lossy means that invalid utf8 values are replaced with � characters. When using other encodings than utf8 or utf8-lossy, the input is first decoded im memory with Ruby.

  • low_memory (Boolean) (defaults to: false) —

    Reduce memory usage at expense of performance.

  • rechunk (Boolean) (defaults to: true) —

    Make sure that all columns are contiguous in memory by aggregating the chunks into a single array.

  • storage_options (Hash) (defaults to: nil) —

    Extra options that make sense for a particular storage connection.

  • skip_rows_after_header (Integer) (defaults to: 0) —

    Skip this number of rows when the header is parsed.

  • row_count_name (String) (defaults to: nil) —

    If not nil, this will insert a row count column with the given name into the DataFrame.

  • row_count_offset (Integer) (defaults to: 0) —

    Offset to start the row_count column (only used if the name is set).

  • eol_char (String) (defaults to: "\n") —

    Single byte end of line character.

  • truncate_ragged_lines (Boolean) (defaults to: false) —

    Truncate lines that are longer than the schema.

Returns:



90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
 
# File 'lib/polars/io/csv.rb', line 90

def read_csv(
  source,
  has_header: true,
  columns: nil,
  new_columns: nil,
  sep: ",",
  comment_char: nil,
  quote_char: '"',
  skip_rows: 0,
  dtypes: nil,
  null_values: nil,
  ignore_errors: false,
  parse_dates: false,
  n_threads: nil,
  infer_schema_length: N_INFER_DEFAULT,
  batch_size: 8192,
  n_rows: nil,
  encoding: "utf8",
  low_memory: false,
  rechunk: true,
  storage_options: nil,
  skip_rows_after_header: 0,
  row_count_name: nil,
  row_count_offset: 0,
  eol_char: "\n",
  truncate_ragged_lines: false
)
  Utils._check_arg_is_1byte("sep", sep, false)
  Utils._check_arg_is_1byte("comment_char", comment_char, false)
  Utils._check_arg_is_1byte("quote_char", quote_char, true)
  Utils._check_arg_is_1byte("eol_char", eol_char, false)

  projection, columns = Utils.handle_projection_columns(columns)

  storage_options ||= {}

  if columns && !has_header
    columns.each do |column|
      if !column.start_with?("column_")
        raise ArgumentError, "Specified column names do not start with \"column_\", but autogenerated header names were requested."
      end
    end
  end

  if projection || new_columns
    raise Todo
  end

  df = nil
  _prepare_file_arg(source) do |data|
    df = _read_csv_impl(
      data,
      has_header: has_header,
      columns: columns || projection,
      sep: sep,
      comment_char: comment_char,
      quote_char: quote_char,
      skip_rows: skip_rows,
      dtypes: dtypes,
      null_values: null_values,
      ignore_errors: ignore_errors,
      parse_dates: parse_dates,
      n_threads: n_threads,
      infer_schema_length: infer_schema_length,
      batch_size: batch_size,
      n_rows: n_rows,
      encoding: encoding == "utf8-lossy" ? encoding : "utf8",
      low_memory: low_memory,
      rechunk: rechunk,
      skip_rows_after_header: skip_rows_after_header,
      row_count_name: row_count_name,
      row_count_offset: row_count_offset,
      eol_char: eol_char,
      truncate_ragged_lines: truncate_ragged_lines
    )
  end

  if new_columns
    Utils._update_columns(df, new_columns)
  else
    df
  end
end

#read_csv_batched(source, has_header: true, columns: nil, new_columns: nil, sep: ",", comment_char: nil, quote_char: '"', skip_rows: 0, dtypes: nil, null_values: nil, missing_utf8_is_empty_string: false, ignore_errors: false, parse_dates: false, n_threads: nil, infer_schema_length: N_INFER_DEFAULT, batch_size: 50_000, n_rows: nil, encoding: "utf8", low_memory: false, rechunk: true, skip_rows_after_header: 0, row_count_name: nil, row_count_offset: 0, eol_char: "\n", raise_if_empty: true, truncate_ragged_lines: false, decimal_comma: false) ⇒ BatchedCsvReader

Read a CSV file in batches.

Upon creation of the BatchedCsvReader, polars will gather statistics and determine the file chunks. After that work will only be done if next_batches is called.

Examples:

reader = Polars.read_csv_batched(
  "./tpch/tables_scale_100/lineitem.tbl", sep: "|", parse_dates: true
)
reader.next_batches(5)

Parameters:

  • source (Object) —

    Path to a file or a file-like object.

  • has_header (Boolean) (defaults to: true) —

    Indicate if the first row of dataset is a header or not. If set to False, column names will be autogenerated in the following format: column_x, with x being an enumeration over every column in the dataset starting at 1.

  • columns (Object) (defaults to: nil) —

    Columns to select. Accepts a list of column indices (starting at zero) or a list of column names.

  • new_columns (Object) (defaults to: nil) —

    Rename columns right after parsing the CSV file. If the given list is shorter than the width of the DataFrame the remaining columns will have their original name.

  • sep (String) (defaults to: ",") —

    Single byte character to use as delimiter in the file.

  • comment_char (String) (defaults to: nil) —

    Single byte character that indicates the start of a comment line, for instance #.

  • quote_char (String) (defaults to: '"') —

    Single byte character used for csv quoting, default = ". Set to nil to turn off special handling and escaping of quotes.

  • skip_rows (Integer) (defaults to: 0) —

    Start reading after skip_rows lines.

  • dtypes (Object) (defaults to: nil) —

    Overwrite dtypes during inference.

  • null_values (Object) (defaults to: nil) —

    Values to interpret as null values. You can provide a:

    • String: All values equal to this string will be null.
    • Array: All values equal to any string in this array will be null.
    • Hash: A hash that maps column name to a null value string.
  • ignore_errors (Boolean) (defaults to: false) —

    Try to keep reading lines if some lines yield errors. First try infer_schema_length: 0 to read all columns as :str to check which values might cause an issue.

  • parse_dates (Boolean) (defaults to: false) —

    Try to automatically parse dates. If this does not succeed, the column remains of data type :str.

  • n_threads (Integer) (defaults to: nil) —

    Number of threads to use in csv parsing. Defaults to the number of physical cpu's of your system.

  • infer_schema_length (Integer) (defaults to: N_INFER_DEFAULT) —

    Maximum number of lines to read to infer schema. If set to 0, all columns will be read as :str. If set to nil, a full table scan will be done (slow).

  • batch_size (Integer) (defaults to: 50_000) —

    Number of lines to read into the buffer at once. Modify this to change performance.

  • n_rows (Integer) (defaults to: nil) —

    Stop reading from CSV file after reading n_rows. During multi-threaded parsing, an upper bound of n_rows rows cannot be guaranteed.

  • encoding ("utf8", "utf8-lossy") (defaults to: "utf8") —

    Lossy means that invalid utf8 values are replaced with � characters. When using other encodings than utf8 or utf8-lossy, the input is first decoded im memory with Ruby. Defaults to utf8.

  • low_memory (Boolean) (defaults to: false) —

    Reduce memory usage at expense of performance.

  • rechunk (Boolean) (defaults to: true) —

    Make sure that all columns are contiguous in memory by aggregating the chunks into a single array.

  • skip_rows_after_header (Integer) (defaults to: 0) —

    Skip this number of rows when the header is parsed.

  • row_count_name (String) (defaults to: nil) —

    If not nil, this will insert a row count column with the given name into the DataFrame.

  • row_count_offset (Integer) (defaults to: 0) —

    Offset to start the row_count column (only used if the name is set).

  • eol_char (String) (defaults to: "\n") —

    Single byte end of line character.

  • truncate_ragged_lines (Boolean) (defaults to: false) —

    Truncate lines that are longer than the schema.

Returns:

  • (BatchedCsvReader) 


400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
 
# File 'lib/polars/io/csv.rb', line 400

def read_csv_batched(
  source,
  has_header: true,
  columns: nil,
  new_columns: nil,
  sep: ",",
  comment_char: nil,
  quote_char: '"',
  skip_rows: 0,
  dtypes: nil,
  null_values: nil,
  missing_utf8_is_empty_string: false,
  ignore_errors: false,
  parse_dates: false,
  n_threads: nil,
  infer_schema_length: N_INFER_DEFAULT,
  batch_size: 50_000,
  n_rows: nil,
  encoding: "utf8",
  low_memory: false,
  rechunk: true,
  skip_rows_after_header: 0,
  row_count_name: nil,
  row_count_offset: 0,
  eol_char: "\n",
  raise_if_empty: true,
  truncate_ragged_lines: false,
  decimal_comma: false
)
  projection, columns = Utils.handle_projection_columns(columns)

  if columns && !has_header
    columns.each do |column|
      if !column.start_with?("column_")
        raise ArgumentError, "Specified column names do not start with \"column_\", but autogenerated header names were requested."
      end
    end
  end

  if projection || new_columns
    raise Todo
  end

  BatchedCsvReader.new(
    source,
    has_header: has_header,
    columns: columns || projection,
    sep: sep,
    comment_char: comment_char,
    quote_char: quote_char,
    skip_rows: skip_rows,
    dtypes: dtypes,
    null_values: null_values,
    missing_utf8_is_empty_string: missing_utf8_is_empty_string,
    ignore_errors: ignore_errors,
    parse_dates: parse_dates,
    n_threads: n_threads,
    infer_schema_length: infer_schema_length,
    batch_size: batch_size,
    n_rows: n_rows,
    encoding: encoding == "utf8-lossy" ? encoding : "utf8",
    low_memory: low_memory,
    rechunk: rechunk,
    skip_rows_after_header: skip_rows_after_header,
    row_count_name: row_count_name,
    row_count_offset: row_count_offset,
    eol_char: eol_char,
    new_columns: new_columns,
    raise_if_empty: raise_if_empty,
    truncate_ragged_lines: truncate_ragged_lines,
    decimal_comma: decimal_comma
  )
end

#read_database(query, schema_overrides: nil) ⇒ DataFrame Also known as: read_sql

Read a SQL query into a DataFrame.

Parameters:

  • query (Object) —

    ActiveRecord::Relation or ActiveRecord::Result.

  • schema_overrides (Hash) (defaults to: nil) —

    A hash mapping column names to dtypes, used to override the schema inferred from the query.

Returns:



12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
 
# File 'lib/polars/io/database.rb', line 12

def read_database(query, schema_overrides: nil)
  if !defined?(ActiveRecord)
    raise Error, "Active Record not available"
  end

  result =
    if query.is_a?(ActiveRecord::Result)
      query
    elsif query.is_a?(ActiveRecord::Relation)
      query.connection_pool.with_connection { |c| c.select_all(query.to_sql) }
    elsif query.is_a?(::String)
      ActiveRecord::Base.connection_pool.with_connection { |c| c.select_all(query) }
    else
      raise ArgumentError, "Expected ActiveRecord::Relation, ActiveRecord::Result, or String"
    end

  data = {}
  schema_overrides = (schema_overrides || {}).transform_keys(&:to_s)

  result.columns.each_with_index do |k, i|
    column_type = result.column_types[i]

    data[k] =
      if column_type
        result.rows.map { |r| column_type.deserialize(r[i]) }
      else
        result.rows.map { |r| r[i] }
      end

    polars_type =
      case column_type&.type
      when :binary
        Binary
      when :boolean
        Boolean
      when :date
        Date
      when :datetime, :timestamp
        Datetime
      when :decimal
        Decimal
      when :float
        Float64
      when :integer
        Int64
      when :string, :text
        String
      when :time
        Time
      # TODO fix issue with null
      # when :json, :jsonb
      #   Struct
      end

    schema_overrides[k] ||= polars_type if polars_type
  end

  DataFrame.new(data, schema_overrides: schema_overrides)
end

#read_ipc(source, columns: nil, n_rows: nil, memory_map: true, storage_options: nil, row_count_name: nil, row_count_offset: 0, rechunk: true) ⇒ DataFrame

Read into a DataFrame from Arrow IPC (Feather v2) file.

Parameters:

  • source (Object) —

    Path to a file or a file-like object.

  • columns (Object) (defaults to: nil) —

    Columns to select. Accepts a list of column indices (starting at zero) or a list of column names.

  • n_rows (Integer) (defaults to: nil) —

    Stop reading from IPC file after reading n_rows.

  • memory_map (Boolean) (defaults to: true) —

    Try to memory map the file. This can greatly improve performance on repeated queries as the OS may cache pages. Only uncompressed IPC files can be memory mapped.

  • storage_options (Hash) (defaults to: nil) —

    Extra options that make sense for a particular storage connection.

  • row_count_name (String) (defaults to: nil) —

    If not nil, this will insert a row count column with give name into the DataFrame.

  • row_count_offset (Integer) (defaults to: 0) —

    Offset to start the row_count column (only use if the name is set).

  • rechunk (Boolean) (defaults to: true) —

    Make sure that all data is contiguous.

Returns:



27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
 
# File 'lib/polars/io/ipc.rb', line 27

def read_ipc(
  source,
  columns: nil,
  n_rows: nil,
  memory_map: true,
  storage_options: nil,
  row_count_name: nil,
  row_count_offset: 0,
  rechunk: true
)
  storage_options ||= {}
  _prepare_file_arg(source, **storage_options) do |data|
    _read_ipc_impl(
      data,
      columns: columns,
      n_rows: n_rows,
      row_count_name: row_count_name,
      row_count_offset: row_count_offset,
      rechunk: rechunk,
      memory_map: memory_map
    )
  end
end

#read_ipc_schema(source) ⇒ Hash

Get a schema of the IPC file without reading data.

Parameters:

  • source (Object) —

    Path to a file or a file-like object.

Returns:

  • (Hash) 


164
165
166
167
168
169
170
 
# File 'lib/polars/io/ipc.rb', line 164

def read_ipc_schema(source)
  if Utils.pathlike?(source)
    source = Utils.normalize_filepath(source)
  end

  Plr.ipc_schema(source)
end

#read_ipc_stream(source, columns: nil, n_rows: nil, storage_options: nil, row_index_name: nil, row_index_offset: 0, rechunk: true) ⇒ DataFrame

Read into a DataFrame from Arrow IPC record batch stream.

See "Streaming format" on https://arrow.apache.org/docs/python/ipc.html.

Parameters:

  • source (Object) —

    Path to a file or a file-like object.

  • columns (Array) (defaults to: nil) —

    Columns to select. Accepts a list of column indices (starting at zero) or a list of column names.

  • n_rows (Integer) (defaults to: nil) —

    Stop reading from IPC stream after reading n_rows.

  • storage_options (Hash) (defaults to: nil) —

    Extra options that make sense for a particular storage connection.

  • row_index_name (String) (defaults to: nil) —

    Insert a row index column with the given name into the DataFrame as the first column. If set to nil (default), no row index column is created.

  • row_index_offset (Integer) (defaults to: 0) —

    Start the row index at this offset. Cannot be negative. Only used if row_index_name is set.

  • rechunk (Boolean) (defaults to: true) —

    Make sure that all data is contiguous.

Returns:



108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
 
# File 'lib/polars/io/ipc.rb', line 108

def read_ipc_stream(
  source,
  columns: nil,
  n_rows: nil,
  storage_options: nil,
  row_index_name: nil,
  row_index_offset: 0,
  rechunk: true
)
  storage_options ||= {}
  _prepare_file_arg(source, **storage_options) do |data|
    _read_ipc_stream_impl(
      data,
      columns: columns,
      n_rows: n_rows,
      row_index_name: row_index_name,
      row_index_offset: row_index_offset,
      rechunk: rechunk
    )
  end
end

#read_json(source, schema: nil, schema_overrides: nil, infer_schema_length: N_INFER_DEFAULT) ⇒ DataFrame

Read into a DataFrame from a JSON file.

Parameters:

  • source (Object) —

    Path to a file or a file-like object.

Returns:



9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
 
# File 'lib/polars/io/json.rb', line 9

def read_json(
  source,
  schema: nil,
  schema_overrides: nil,
  infer_schema_length: N_INFER_DEFAULT
)
  if Utils.pathlike?(source)
    source = Utils.normalize_filepath(source)
  end

  rbdf =
    RbDataFrame.read_json(
      source,
      infer_schema_length,
      schema,
      schema_overrides
    )
  Utils.wrap_df(rbdf)
end

#read_ndjson(source, schema: nil, schema_overrides: nil, ignore_errors: false) ⇒ DataFrame

Read into a DataFrame from a newline delimited JSON file.

Parameters:

  • source (Object) —

    Path to a file or a file-like object.

Returns:



9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
 
# File 'lib/polars/io/ndjson.rb', line 9

def read_ndjson(
  source,
  schema: nil,
  schema_overrides: nil,
  ignore_errors: false
)
  if Utils.pathlike?(source)
    source = Utils.normalize_filepath(source)
  end

  rbdf =
    RbDataFrame.read_ndjson(
      source,
      ignore_errors,
      schema,
      schema_overrides
    )
  Utils.wrap_df(rbdf)
end

#read_parquet(source, columns: nil, n_rows: nil, row_count_name: nil, row_count_offset: 0, parallel: "auto", use_statistics: true, hive_partitioning: nil, glob: true, schema: nil, hive_schema: nil, try_parse_hive_dates: true, rechunk: false, low_memory: false, storage_options: nil, credential_provider: nil, retries: 2, include_file_paths: nil, allow_missing_columns: false) ⇒ DataFrame

Read into a DataFrame from a parquet file.

Parameters:

  • source (Object) —

    Path to a file or a file-like object.

  • columns (Object) (defaults to: nil) —

    Columns to select. Accepts a list of column indices (starting at zero) or a list of column names.

  • n_rows (Integer) (defaults to: nil) —

    Stop reading from parquet file after reading n_rows.

  • row_count_name (String) (defaults to: nil) —

    If not nil, this will insert a row count column with give name into the DataFrame.

  • row_count_offset (Integer) (defaults to: 0) —

    Offset to start the row_count column (only use if the name is set).

  • parallel ("auto", "columns", "row_groups", "none") (defaults to: "auto") —

    This determines the direction of parallelism. 'auto' will try to determine the optimal direction.

  • use_statistics (Boolean) (defaults to: true) —

    Use statistics in the parquet to determine if pages can be skipped from reading.

  • hive_partitioning (Boolean) (defaults to: nil) —

    Infer statistics and schema from hive partitioned URL and use them to prune reads.

  • glob (Boolean) (defaults to: true) —

    Expand path given via globbing rules.

  • schema (Object) (defaults to: nil) —

    Specify the datatypes of the columns. The datatypes must match the datatypes in the file(s). If there are extra columns that are not in the file(s), consider also enabling allow_missing_columns.

  • hive_schema (Object) (defaults to: nil) —

    The column names and data types of the columns by which the data is partitioned. If set to nil (default), the schema of the Hive partitions is inferred.

  • try_parse_hive_dates (Boolean) (defaults to: true) —

    Whether to try parsing hive values as date/datetime types.

  • rechunk (Boolean) (defaults to: false) —

    In case of reading multiple files via a glob pattern rechunk the final DataFrame into contiguous memory chunks.

  • low_memory (Boolean) (defaults to: false) —

    Reduce memory pressure at the expense of performance.

  • storage_options (Hash) (defaults to: nil) —

    Extra options that make sense for a particular storage connection.

  • credential_provider (Object) (defaults to: nil) —

    Provide a function that can be called to provide cloud storage credentials. The function is expected to return a dictionary of credential keys along with an optional credential expiry time.

  • retries (Integer) (defaults to: 2) —

    Number of retries if accessing a cloud instance fails.

  • include_file_paths (String) (defaults to: nil) —

    Include the path of the source file(s) as a column with this name.

Returns:



54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
 
# File 'lib/polars/io/parquet.rb', line 54

def read_parquet(
  source,
  columns: nil,
  n_rows: nil,
  row_count_name: nil,
  row_count_offset: 0,
  parallel: "auto",
  use_statistics: true,
  hive_partitioning: nil,
  glob: true,
  schema: nil,
  hive_schema: nil,
  try_parse_hive_dates: true,
  rechunk: false,
  low_memory: false,
  storage_options: nil,
  credential_provider: nil,
  retries: 2,
  include_file_paths: nil,
  allow_missing_columns: false
)
  lf =
    scan_parquet(
      source,
      n_rows: n_rows,
      row_count_name: row_count_name,
      row_count_offset: row_count_offset,
      parallel: parallel,
      use_statistics: use_statistics,
      hive_partitioning: hive_partitioning,
      schema: schema,
      hive_schema: hive_schema,
      try_parse_hive_dates: try_parse_hive_dates,
      rechunk: rechunk,
      low_memory: low_memory,
      cache: false,
      storage_options: storage_options,
      credential_provider: credential_provider,
      retries: retries,
      glob: glob,
      include_file_paths: include_file_paths,
      allow_missing_columns: allow_missing_columns
    )

  if !columns.nil?
    if Utils.is_int_sequence(columns)
      lf = lf.select(F.nth(columns))
    else
      lf = lf.select(columns)
    end
  end

  lf.collect
end

#read_parquet_schema(source) ⇒ Hash

Get a schema of the Parquet file without reading data.

Parameters:

  • source (Object) —

    Path to a file or a file-like object.

Returns:

  • (Hash) 


115
116
117
118
119
120
121
 
# File 'lib/polars/io/parquet.rb', line 115

def read_parquet_schema(source)
  if Utils.pathlike?(source)
    source = Utils.normalize_filepath(source)
  end

  Plr.parquet_schema(source)
end

#scan_csv(source, has_header: true, sep: ",", comment_char: nil, quote_char: '"', skip_rows: 0, dtypes: nil, null_values: nil, missing_utf8_is_empty_string: false, ignore_errors: false, cache: true, with_column_names: nil, infer_schema_length: N_INFER_DEFAULT, n_rows: nil, encoding: "utf8", low_memory: false, rechunk: true, skip_rows_after_header: 0, row_count_name: nil, row_count_offset: 0, parse_dates: false, eol_char: "\n", raise_if_empty: true, truncate_ragged_lines: false, decimal_comma: false, glob: true) ⇒ LazyFrame

Lazily read from a CSV file or multiple files via glob patterns.

This allows the query optimizer to push down predicates and projections to the scan level, thereby potentially reducing memory overhead.

Parameters:

  • source (Object) —

    Path to a file.

  • has_header (Boolean) (defaults to: true) —

    Indicate if the first row of dataset is a header or not. If set to false, column names will be autogenerated in the following format: column_x, with x being an enumeration over every column in the dataset starting at 1.

  • sep (String) (defaults to: ",") —

    Single byte character to use as delimiter in the file.

  • comment_char (String) (defaults to: nil) —

    Single byte character that indicates the start of a comment line, for instance #.

  • quote_char (String) (defaults to: '"') —

    Single byte character used for csv quoting. Set to None to turn off special handling and escaping of quotes.

  • skip_rows (Integer) (defaults to: 0) —

    Start reading after skip_rows lines. The header will be parsed at this offset.

  • dtypes (Object) (defaults to: nil) —

    Overwrite dtypes during inference.

  • null_values (Object) (defaults to: nil) —

    Values to interpret as null values. You can provide a:

    • String: All values equal to this string will be null.
    • Array: All values equal to any string in this array will be null.
    • Hash: A hash that maps column name to a null value string.
  • ignore_errors (Boolean) (defaults to: false) —

    Try to keep reading lines if some lines yield errors. First try infer_schema_length: 0 to read all columns as :str to check which values might cause an issue.

  • cache (Boolean) (defaults to: true) —

    Cache the result after reading.

  • with_column_names (Object) (defaults to: nil) —

    Apply a function over the column names. This can be used to update a schema just in time, thus before scanning.

  • infer_schema_length (Integer) (defaults to: N_INFER_DEFAULT) —

    Maximum number of lines to read to infer schema. If set to 0, all columns will be read as :str. If set to nil, a full table scan will be done (slow).

  • n_rows (Integer) (defaults to: nil) —

    Stop reading from CSV file after reading n_rows.

  • encoding ("utf8", "utf8-lossy") (defaults to: "utf8") —

    Lossy means that invalid utf8 values are replaced with � characters.

  • low_memory (Boolean) (defaults to: false) —

    Reduce memory usage in expense of performance.

  • rechunk (Boolean) (defaults to: true) —

    Reallocate to contiguous memory when all chunks/ files are parsed.

  • skip_rows_after_header (Integer) (defaults to: 0) —

    Skip this number of rows when the header is parsed.

  • row_count_name (String) (defaults to: nil) —

    If not nil, this will insert a row count column with the given name into the DataFrame.

  • row_count_offset (Integer) (defaults to: 0) —

    Offset to start the row_count column (only used if the name is set).

  • parse_dates (Boolean) (defaults to: false) —

    Try to automatically parse dates. If this does not succeed, the column remains of data type :str.

  • eol_char (String) (defaults to: "\n") —

    Single byte end of line character.

  • truncate_ragged_lines (Boolean) (defaults to: false) —

    Truncate lines that are longer than the schema.

Returns:



545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
 
# File 'lib/polars/io/csv.rb', line 545

def scan_csv(
  source,
  has_header: true,
  sep: ",",
  comment_char: nil,
  quote_char: '"',
  skip_rows: 0,
  dtypes: nil,
  null_values: nil,
  missing_utf8_is_empty_string: false,
  ignore_errors: false,
  cache: true,
  with_column_names: nil,
  infer_schema_length: N_INFER_DEFAULT,
  n_rows: nil,
  encoding: "utf8",
  low_memory: false,
  rechunk: true,
  skip_rows_after_header: 0,
  row_count_name: nil,
  row_count_offset: 0,
  parse_dates: false,
  eol_char: "\n",
  raise_if_empty: true,
  truncate_ragged_lines: false,
  decimal_comma: false,
  glob: true
)
  Utils._check_arg_is_1byte("sep", sep, false)
  Utils._check_arg_is_1byte("comment_char", comment_char, false)
  Utils._check_arg_is_1byte("quote_char", quote_char, true)

  if Utils.pathlike?(source)
    source = Utils.normalize_filepath(source)
  end

  _scan_csv_impl(
    source,
    has_header: has_header,
    sep: sep,
    comment_char: comment_char,
    quote_char: quote_char,
    skip_rows: skip_rows,
    dtypes: dtypes,
    null_values: null_values,
    ignore_errors: ignore_errors,
    cache: cache,
    with_column_names: with_column_names,
    infer_schema_length: infer_schema_length,
    n_rows: n_rows,
    low_memory: low_memory,
    rechunk: rechunk,
    skip_rows_after_header: skip_rows_after_header,
    encoding: encoding,
    row_count_name: row_count_name,
    row_count_offset: row_count_offset,
    parse_dates: parse_dates,
    eol_char: eol_char,
    truncate_ragged_lines: truncate_ragged_lines
  )
end

#scan_ipc(source, n_rows: nil, cache: true, rechunk: true, row_count_name: nil, row_count_offset: 0, storage_options: nil, hive_partitioning: nil, hive_schema: nil, try_parse_hive_dates: true, include_file_paths: nil) ⇒ LazyFrame

Lazily read from an Arrow IPC (Feather v2) file or multiple files via glob patterns.

This allows the query optimizer to push down predicates and projections to the scan level, thereby potentially reducing memory overhead.

Parameters:

  • source (String) —

    Path to a IPC file.

  • n_rows (Integer) (defaults to: nil) —

    Stop reading from IPC file after reading n_rows.

  • cache (Boolean) (defaults to: true) —

    Cache the result after reading.

  • rechunk (Boolean) (defaults to: true) —

    Reallocate to contiguous memory when all chunks/ files are parsed.

  • row_count_name (String) (defaults to: nil) —

    If not nil, this will insert a row count column with give name into the DataFrame.

  • row_count_offset (Integer) (defaults to: 0) —

    Offset to start the row_count column (only use if the name is set).

  • storage_options (Hash) (defaults to: nil) —

    Extra options that make sense for a particular storage connection.

  • hive_partitioning (Boolean) (defaults to: nil) —

    Infer statistics and schema from Hive partitioned URL and use them to prune reads. This is unset by default (i.e. nil), meaning it is automatically enabled when a single directory is passed, and otherwise disabled.

  • hive_schema (Hash) (defaults to: nil) —

    The column names and data types of the columns by which the data is partitioned. If set to nil (default), the schema of the Hive partitions is inferred.

  • try_parse_hive_dates (Boolean) (defaults to: true) —

    Whether to try parsing hive values as date/datetime types.

  • include_file_paths (String) (defaults to: nil) —

    Include the path of the source file(s) as a column with this name.

Returns:



206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
 
# File 'lib/polars/io/ipc.rb', line 206

def scan_ipc(
  source,
  n_rows: nil,
  cache: true,
  rechunk: true,
  row_count_name: nil,
  row_count_offset: 0,
  storage_options: nil,
  hive_partitioning: nil,
  hive_schema: nil,
  try_parse_hive_dates: true,
  include_file_paths: nil
)
  _scan_ipc_impl(
    source,
    n_rows: n_rows,
    cache: cache,
    rechunk: rechunk,
    row_count_name: row_count_name,
    row_count_offset: row_count_offset,
    storage_options: storage_options,
    hive_partitioning: hive_partitioning,
    hive_schema: hive_schema,
    try_parse_hive_dates: try_parse_hive_dates,
    include_file_paths: include_file_paths
  )
end

#scan_ndjson(source, infer_schema_length: N_INFER_DEFAULT, batch_size: 1024, n_rows: nil, low_memory: false, rechunk: true, row_count_name: nil, row_count_offset: 0) ⇒ LazyFrame

Lazily read from a newline delimited JSON file.

This allows the query optimizer to push down predicates and projections to the scan level, thereby potentially reducing memory overhead.

Parameters:

  • source (String) —

    Path to a file.

  • infer_schema_length (Integer) (defaults to: N_INFER_DEFAULT) —

    Infer the schema length from the first infer_schema_length rows.

  • batch_size (Integer) (defaults to: 1024) —

    Number of rows to read in each batch.

  • n_rows (Integer) (defaults to: nil) —

    Stop reading from JSON file after reading n_rows.

  • low_memory (Boolean) (defaults to: false) —

    Reduce memory pressure at the expense of performance.

  • rechunk (Boolean) (defaults to: true) —

    Reallocate to contiguous memory when all chunks/ files are parsed.

  • row_count_name (String) (defaults to: nil) —

    If not nil, this will insert a row count column with give name into the DataFrame.

  • row_count_offset (Integer) (defaults to: 0) —

    Offset to start the row_count column (only use if the name is set).

Returns:



53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
 
# File 'lib/polars/io/ndjson.rb', line 53

def scan_ndjson(
  source,
  infer_schema_length: N_INFER_DEFAULT,
  batch_size: 1024,
  n_rows: nil,
  low_memory: false,
  rechunk: true,
  row_count_name: nil,
  row_count_offset: 0
)
  sources = []
  if Utils.pathlike?(source)
    source = Utils.normalize_filepath(source)
  elsif source.is_a?(::Array)
    if Utils.is_path_or_str_sequence(source)
      sources = source.map { |s| Utils.normalize_filepath(s) }
    else
      sources = source
    end

    source = nil
  end

  rblf =
    RbLazyFrame.new_from_ndjson(
      source,
      sources,
      infer_schema_length,
      batch_size,
      n_rows,
      low_memory,
      rechunk,
      Utils.parse_row_index_args(row_count_name, row_count_offset)
    )
  Utils.wrap_ldf(rblf)
end

#scan_parquet(source, n_rows: nil, row_count_name: nil, row_count_offset: 0, parallel: "auto", use_statistics: true, hive_partitioning: nil, glob: true, schema: nil, hive_schema: nil, try_parse_hive_dates: true, rechunk: false, low_memory: false, cache: true, storage_options: nil, credential_provider: nil, retries: 2, include_file_paths: nil, allow_missing_columns: false) ⇒ LazyFrame

Lazily read from a parquet file or multiple files via glob patterns.

This allows the query optimizer to push down predicates and projections to the scan level, thereby potentially reducing memory overhead.

Parameters:

  • source (Object) —

    Path to a file or a file-like object.

  • n_rows (Integer) (defaults to: nil) —

    Stop reading from parquet file after reading n_rows.

  • row_count_name (String) (defaults to: nil) —

    If not nil, this will insert a row count column with give name into the DataFrame.

  • row_count_offset (Integer) (defaults to: 0) —

    Offset to start the row_count column (only use if the name is set).

  • parallel ("auto", "columns", "row_groups", "none") (defaults to: "auto") —

    This determines the direction of parallelism. 'auto' will try to determine the optimal direction.

  • use_statistics (Boolean) (defaults to: true) —

    Use statistics in the parquet to determine if pages can be skipped from reading.

  • hive_partitioning (Boolean) (defaults to: nil) —

    Infer statistics and schema from hive partitioned URL and use them to prune reads.

  • glob (Boolean) (defaults to: true) —

    Expand path given via globbing rules.

  • schema (Object) (defaults to: nil) —

    Specify the datatypes of the columns. The datatypes must match the datatypes in the file(s). If there are extra columns that are not in the file(s), consider also enabling allow_missing_columns.

  • hive_schema (Object) (defaults to: nil) —

    The column names and data types of the columns by which the data is partitioned. If set to nil (default), the schema of the Hive partitions is inferred.

  • try_parse_hive_dates (Boolean) (defaults to: true) —

    Whether to try parsing hive values as date/datetime types.

  • rechunk (Boolean) (defaults to: false) —

    In case of reading multiple files via a glob pattern rechunk the final DataFrame into contiguous memory chunks.

  • low_memory (Boolean) (defaults to: false) —

    Reduce memory pressure at the expense of performance.

  • cache (Boolean) (defaults to: true) —

    Cache the result after reading.

  • storage_options (Hash) (defaults to: nil) —

    Extra options that make sense for a particular storage connection.

  • credential_provider (Object) (defaults to: nil) —

    Provide a function that can be called to provide cloud storage credentials. The function is expected to return a dictionary of credential keys along with an optional credential expiry time.

  • retries (Integer) (defaults to: 2) —

    Number of retries if accessing a cloud instance fails.

  • include_file_paths (String) (defaults to: nil) —

    Include the path of the source file(s) as a column with this name.

Returns:



176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
 
# File 'lib/polars/io/parquet.rb', line 176

def scan_parquet(
  source,
  n_rows: nil,
  row_count_name: nil,
  row_count_offset: 0,
  parallel: "auto",
  use_statistics: true,
  hive_partitioning: nil,
  glob: true,
  schema: nil,
  hive_schema: nil,
  try_parse_hive_dates: true,
  rechunk: false,
  low_memory: false,
  cache: true,
  storage_options: nil,
  credential_provider: nil,
  retries: 2,
  include_file_paths: nil,
  allow_missing_columns: false
)
  if Utils.pathlike?(source)
    source = Utils.normalize_filepath(source, check_not_directory: false)
  elsif Utils.is_path_or_str_sequence(source)
    source = source.map { |s| Utils.normalize_filepath(s, check_not_directory: false) }
  end

  if credential_provider
    raise Todo
  end

  _scan_parquet_impl(
    source,
    n_rows: n_rows,
    cache: cache,
    parallel: parallel,
    rechunk: rechunk,
    row_index_name: row_count_name,
    row_index_offset: row_count_offset,
    storage_options: storage_options,
    credential_provider: credential_provider,
    low_memory: low_memory,
    use_statistics: use_statistics,
    hive_partitioning: hive_partitioning,
    schema: schema,
    hive_schema: hive_schema,
    try_parse_hive_dates: try_parse_hive_dates,
    retries: retries,
    glob: glob,
    include_file_paths: include_file_paths,
    allow_missing_columns: allow_missing_columns
  )
end