Impala

One goal of Ibis is to provide an integrated Python API for an Impala cluster without requiring you to switch back and forth between Python code and the Impala shell (where one would be using a mix of DDL and SQL statements).

If you find an Impala task that you cannot perform with Ibis, please get in touch on the GitHub issue tracker.

While interoperability between the Hadoop / Spark ecosystems and pandas / the PyData stack is overall poor (but improving), we also show some ways that you can use pandas with Ibis and Impala.

Impala Quickstart

Install dependencies for Ibis’s Impala dialect:

pip install 'ibis-framework[impala]'

To create an Ibis client, you must first connect your services and assemble the client using ibis.impala.connect():

import ibis

hdfs = ibis.impala.hdfs_connect(host='impala', port=50070)
con = ibis.impala.connect(
    host='impala', database='ibis_testing', hdfs_client=hdfs
)

Both method calls can take auth_mechanism='GSSAPI' or auth_mechanism='LDAP' to connect to Kerberos clusters. Depending on your cluster setup, this may also include SSL. See the API reference for more, along with the Impala shell reference, as the connection semantics are identical.

API

These methods are available on the Impala client object after connecting to your HDFS cluster (ibis.impala.hdfs_connect) and connecting to Impala with ibis.impala.connect. See Impala for a tutorial on using this backend.

Backend.connect([host, port, database, ...])

Create an ImpalaClient for use with Ibis.

ImpalaClient.close()

Close Impala connection and drop any temporary objects

ImpalaClient.database([name])

Database methods

ImpalaClient.set_database(name)

ImpalaClient.create_database(name[, path, force])

Create a new Impala database

ImpalaClient.drop_database(name[, force])

Drop an Impala database.

ImpalaClient.list_databases([like])

ImpalaClient.exists_database(name)

ImpalaDatabase.create_table(table_name[, obj])

Dispatch to ImpalaClient.create_table.

ImpalaDatabase.drop([force])

Drop the database.

ImpalaDatabase.table(name)

Return a table expression referencing a table in this database.

Table methods

The ImpalaClient object itself has many helper utility methods. You’ll find the most methods on ImpalaTable.

ImpalaClient.database([name])

ImpalaClient.table(name[, database])

Create a table expression.

ImpalaClient.sql(query)

Convert a SQL query to an Ibis table expression.

ImpalaClient.raw_sql(query[, results])

Execute a given query string.

ImpalaClient.list_tables([like, database])

ImpalaClient.exists_table(name[, database])

Determine if the indicated table or view exists

ImpalaClient.drop_table(table_name[, ...])

Drop an Impala table

ImpalaClient.create_table(table_name[, obj, ...])

Create a new table in Impala using an Ibis table expression.

ImpalaClient.insert(table_name[, obj, ...])

Insert into existing table.

ImpalaClient.invalidate_metadata([name, ...])

Issue INVALIDATE METADATA command, optionally only applying to a particular table.

ImpalaClient.truncate_table(table_name[, ...])

Delete all rows from, but do not drop, an existing table

ImpalaClient.get_schema(table_name[, database])

Return a Schema object for the indicated table and database

ImpalaClient.cache_table(table_name[, ...])

Caches a table in cluster memory in the given pool.

ImpalaClient.load_data(table_name, path[, ...])

Wraps the LOAD DATA DDL statement.

ImpalaClient.get_options()

Return current query options for the Impala session

ImpalaClient.set_options(options)

ImpalaClient.set_compression_codec(codec)

Parameters

The best way to interact with a single table is through the ImpalaTable object you get back from ImpalaClient.table.

ImpalaTable.add_partition(spec[, location])

Add a new table partition, creating any new directories in HDFS if necessary.

ImpalaTable.alter([location, format, ...])

Change setting and parameters of the table.

ImpalaTable.alter_partition(spec[, ...])

Change setting and parameters of an existing partition

ImpalaTable.column_stats()

Return results of SHOW COLUMN STATS as a pandas DataFrame

ImpalaTable.compute_stats([incremental])

Invoke Impala COMPUTE STATS command to compute column, table, and partition statistics.

ImpalaTable.describe_formatted()

Return parsed results of DESCRIBE FORMATTED statement

ImpalaTable.drop()

Drop the table from the database

ImpalaTable.drop_partition(spec)

Drop an existing table partition

ImpalaTable.files()

Return results of SHOW FILES statement

ImpalaTable.insert([obj, overwrite, ...])

Insert into Impala table.

ImpalaTable.invalidate_metadata()

ImpalaTable.is_partitioned

True if the table is partitioned

ImpalaTable.load_data(path[, overwrite, ...])

Wraps the LOAD DATA DDL statement.

ImpalaTable.metadata()

Return parsed results of DESCRIBE FORMATTED statement

ImpalaTable.partition_schema()

For partitioned tables, return the schema (names and types) for the partition columns

ImpalaTable.partitions()

Return a pandas.DataFrame giving information about this table's partitions.

ImpalaTable.refresh()

ImpalaTable.rename(new_name[, database])

Rename table inside Impala.

ImpalaTable.schema()

Get the schema for this table (if one is known)

ImpalaTable.stats()

Return results of SHOW TABLE STATS as a DataFrame.

Creating views is also possible:

ImpalaClient.create_view(name, expr[, database])

Create an Impala view from a table expression

ImpalaClient.drop_view(name[, database, force])

Drop an Impala view

ImpalaClient.drop_table_or_view(name[, ...])

Attempt to drop a relation that may be a view or table

Accessing data formats in HDFS

ImpalaClient.avro_file(hdfs_dir, avro_schema)

Create a (possibly temporary) table to read a collection of Avro data.

ImpalaClient.delimited_file(hdfs_dir, schema)

Interpret delimited text files (CSV / TSV / etc.) as an Ibis table.

ImpalaClient.parquet_file(hdfs_dir[, ...])

Make indicated parquet file in HDFS available as an Ibis table.

Executing expressions

ImpalaClient.execute(expr[, params, limit])

Compile and execute the given Ibis expression.

ImpalaClient.disable_codegen([disabled])

Turn off or on LLVM codegen in Impala query execution

HDFS

Client objects have an hdfs attribute you can use to interact directly with HDFS.

hdfs_connect([host, port, protocol, ...])

Connect to HDFS.

HDFS.ls(hdfs_path[, status])

Return contents of directory.

HDFS.chmod(hdfs_path, permissions)

Change permissions of a file of directory.

HDFS.chown(hdfs_path[, owner, group])

Change owner (and/or group) of a file or directory.

HDFS.get(hdfs_path[, local_path, overwrite])

Download remote file or directory to the local filesystem.

HDFS.head(hdfs_path[, nbytes, offset])

Retrieve the requested number of bytes from a file.

HDFS.put(hdfs_path, resource[, overwrite, ...])

Write file or directory to HDFS.

HDFS.put_tarfile(hdfs_path, local_path[, ...])

Write contents of tar archive to HDFS.

HDFS.rm(path)

Delete a single file.

HDFS.rmdir(path)

Delete a directory and all its contents.

HDFS.size(hdfs_path)

Return total size of file or directory.

HDFS.status(path)

Check if the status of the path.

The Impala client object

To use Ibis with Impala, you first must connect to a cluster using the ibis.impala.connect() function, optionally supplying an HDFS connection:

import ibis

hdfs = ibis.impala.hdfs_connect(host=webhdfs_host, port=webhdfs_port)
client = ibis.impala.connect(
    host=impala_host, port=impala_port, hdfs_client=hdfs
)

All examples here use the following block of code to connect to impala using docker:

import ibis
host = 'impala'
hdfs = ibis.impala.hdfs_connect(host=host)
client = ibis.impala.connect(host=host, hdfs_client=hdfs)

You can accomplish many tasks directly through the client object, but we additionally provide APIs to streamline tasks involving a single Impala table or database.

Database and Table objects

ImpalaClient.database([name])

ImpalaClient.table(name[, database])

Create a table expression.

The client’s table method allows you to create an Ibis table expression referencing a physical Impala table:

table = client.table('functional_alltypes', database='ibis_testing')

While you can get by fine with only table and client objects, Ibis has a notion of a database object that simplifies interactions with a single Impala database. It also gives you IPython tab completion of table names (that are valid Python variable names):

>>> db = client.database('ibis_testing')
>>> db
ImpalaDatabase('ibis_testing')

>>> table = db.functional_alltypes
>>> db.list_tables()
['alltypes',
 'functional_alltypes',
 'tpch_customer',
 'tpch_lineitem',
 'tpch_nation',
 'tpch_orders',
 'tpch_part',
 'tpch_partsupp',
 'tpch_region',
 'tpch_region_avro',
 'tpch_supplier']

ImpalaTable is a Python subclass of the more general Ibis TableExpr that has additional Impala-specific methods. So you can use it interchangeably with any code expecting a TableExpr.

Like all table expressions in Ibis, ImpalaTable has a schema method you can use to examine its schema:

ImpalaTable.schema()

Get the schema for this table (if one is known)

While the client has a drop_table method you can use to drop tables, the table itself has a method drop that you can use:

table.drop()

Expression execution

Ibis expressions have an execute method with compiles and runs the expressions on Impala or whichever backend is being referenced.

For example:

>>> fa = db.functional_alltypes
>>> expr = fa.double_col.sum()
>>> expr.execute()
331785.00000000006

For longer-running queries, Ibis will attempt to cancel the query in progress if an interrupt is received.

Creating tables

There are several ways to create new Impala tables:

  • From an Ibis table expression

  • Empty, from a declared schema

  • Empty and partitioned

In all cases, you should use the create_table method either on the top-level client connection or a database object.

ImpalaClient.create_table(table_name[, obj, ...])

Create a new table in Impala using an Ibis table expression.

ImpalaDatabase.create_table(table_name[, obj])

Dispatch to ImpalaClient.create_table.

Creating tables from a table expression

If you pass an Ibis expression to create_table, Ibis issues a CREATE TABLE .. AS SELECT (CTAS) statement:

>>> table = db.table('functional_alltypes')
>>> expr = table.group_by('string_col').size()
>>> db.create_table('string_freqs', expr, format='parquet')

>>> freqs = db.table('string_freqs')
>>> freqs.execute()
  string_col  count
0          9    730
1          3    730
2          6    730
3          4    730
4          1    730
5          8    730
6          2    730
7          7    730
8          5    730
9          0    730

>>> files = freqs.files()
>>> files
                                                Path  Size Partition
0  hdfs://impala:8020/user/hive/warehouse/ibis_te...  584B

>>> freqs.drop()

You can also choose to create an empty table and use insert (see below).

Creating an empty table

To create an empty table, you must declare an Ibis schema that will be translated to the appopriate Impala schema and data types.

As Ibis types are simplified compared with Impala types, this may expand in the future to include a more fine-grained schema declaration.

You can use the create_table method either on a database or client object.

schema = ibis.schema([('foo', 'string'),
                      ('year', 'int32'),
                      ('month', 'int16')])
name = 'new_table'
db.create_table(name, schema=schema)

By default, this stores the data files in the database default location. You can force a particular path with the location option.

from getpass import getuser
schema = ibis.schema([('foo', 'string'),
                      ('year', 'int32'),
                      ('month', 'int16')])
name = 'new_table'
location = '/home/{}/new-table-data'.format(getuser())
db.create_table(name, schema=schema, location=location)

If the schema matches a known table schema, you can always use the schema method to get a schema object:

>>> t = db.table('functional_alltypes')
>>> t.schema()
ibis.Schema {
  id               int32
  bool_col         boolean
  tinyint_col      int8
  smallint_col     int16
  int_col          int32
  bigint_col       int64
  float_col        float32
  double_col       float64
  date_string_col  string
  string_col       string
  timestamp_col    timestamp
  year             int32
  month            int32
}

Creating a partitioned table

To create an empty partitioned table, include a list of columns to be used as the partition keys.

schema = ibis.schema([('foo', 'string'),
                      ('year', 'int32'),
                      ('month', 'int16')])
name = 'new_table'
db.create_table(name, schema=schema, partition=['year', 'month'])

Partitioned tables

Ibis enables you to manage partitioned tables in various ways. Since each partition behaves as its own “subtable” sharing a common schema, each partition can have its own file format, directory path, serialization properties, and so forth.

There are a handful of table methods for adding and removing partitions and getting information about the partition schema and any existing partition data:

ImpalaTable.add_partition(spec[, location])

Add a new table partition, creating any new directories in HDFS if necessary.

ImpalaTable.drop_partition(spec)

Drop an existing table partition

ImpalaTable.is_partitioned

True if the table is partitioned

ImpalaTable.partition_schema()

For partitioned tables, return the schema (names and types) for the partition columns

ImpalaTable.partitions()

Return a pandas.DataFrame giving information about this table's partitions.

To address a specific partition in any method that is partition specific, you can either use a dict with the partition key names and values, or pass a list of the partition values:

schema = ibis.schema([('foo', 'string'),
                      ('year', 'int32'),
                      ('month', 'int16')])
name = 'new_table'
db.create_table(name, schema=schema, partition=['year', 'month'])

table = db.table(name)

table.add_partition({'year': 2007, 'month', 4})
table.add_partition([2007, 5])
table.add_partition([2007, 6])

table.drop_partition([2007, 6])

We’ll cover partition metadata management and data loading below.

Inserting data into tables

If the schemas are compatible, you can insert into a table directly from an Ibis table expression:

>>> t = db.functional_alltypes
>>> db.create_table('insert_test', schema=t.schema())
>>> target = db.table('insert_test')

>>> target.insert(t[:3])
>>> target.insert(t[:3])
>>> target.insert(t[:3])

>>> target.execute()
     id  bool_col  tinyint_col  ...           timestamp_col  year  month
0  5770      True            0  ... 2010-08-01 00:00:00.000  2010      8
1  5771     False            1  ... 2010-08-01 00:01:00.000  2010      8
2  5772      True            2  ... 2010-08-01 00:02:00.100  2010      8
3  5770      True            0  ... 2010-08-01 00:00:00.000  2010      8
4  5771     False            1  ... 2010-08-01 00:01:00.000  2010      8
5  5772      True            2  ... 2010-08-01 00:02:00.100  2010      8
6  5770      True            0  ... 2010-08-01 00:00:00.000  2010      8
7  5771     False            1  ... 2010-08-01 00:01:00.000  2010      8
8  5772      True            2  ... 2010-08-01 00:02:00.100  2010      8

[9 rows x 13 columns]

>>> target.drop()

If the table is partitioned, you must indicate the partition you are inserting into:

part = {'year': 2007, 'month': 4}
table.insert(expr, partition=part)

Managing table metadata

Ibis has functions that wrap many of the DDL commands for Impala table metadata.

Detailed table metadata: DESCRIBE FORMATTED

To get a handy wrangled version of DESCRIBE FORMATTED use the metadata method.

ImpalaTable.metadata()

Return parsed results of DESCRIBE FORMATTED statement

The TableMetadata object that is returned has a nicer console output and many attributes set that you can explore in IPython:

>>> t = client.table('ibis_testing.functional_alltypes')
>>> meta = t.metadata()
>>> meta
<class 'ibis.backends.impala.metadata.TableMetadata'>
{'info': {'CreateTime': datetime.datetime(2021, 1, 14, 21, 23, 8),
          'Database': 'ibis_testing',
          'LastAccessTime': 'UNKNOWN',
          'Location': 'hdfs://impala:8020/__ibis/ibis-testing-data/parquet/functional_alltypes',
          'Owner': 'root',
          'Protect Mode': 'None',
          'Retention': 0,
          'Table Parameters': {'COLUMN_STATS_ACCURATE': False,
                               'EXTERNAL': True,
                               'STATS_GENERATED_VIA_STATS_TASK': True,
                               'numFiles': 3,
                               'numRows': 7300,
                               'rawDataSize': '-1',
                               'totalSize': 106278,
                               'transient_lastDdlTime': datetime.datetime(2021, 1, 14, 21, 23, 17)},
          'Table Type': 'EXTERNAL_TABLE'},
 'schema': [('id', 'int'),
            ('bool_col', 'boolean'),
            ('tinyint_col', 'tinyint'),
            ('smallint_col', 'smallint'),
            ('int_col', 'int'),
            ('bigint_col', 'bigint'),
            ('float_col', 'float'),
            ('double_col', 'double'),
            ('date_string_col', 'string'),
            ('string_col', 'string'),
            ('timestamp_col', 'timestamp'),
            ('year', 'int'),
            ('month', 'int')],
 'storage info': {'Bucket Columns': '[]',
                  'Compressed': False,
                  'InputFormat': 'org.apache.hadoop.hive.ql.io.parquet.MapredParquetInputFormat',
                  'Num Buckets': 0,
                  'OutputFormat': 'org.apache.hadoop.hive.ql.io.parquet.MapredParquetOutputFormat',
                  'SerDe Library': 'org.apache.hadoop.hive.ql.io.parquet.serde.ParquetHiveSerDe',
                  'Sort Columns': '[]'}}

>>> meta.location
'hdfs://impala:8020/__ibis/ibis-testing-data/parquet/functional_alltypes'

>>> meta.create_time
datetime.datetime(2021, 1, 14, 21, 23, 8)

The files function is also available to see all of the physical HDFS data files backing a table:

ImpalaTable.files()

Return results of SHOW FILES statement

>>> ss = c.table('tpcds_parquet.store_sales')

>>> ss.files()[:5]
                                                path      size  \
0  hdfs://localhost:20500/test-warehouse/tpcds.st...  160.61KB
1  hdfs://localhost:20500/test-warehouse/tpcds.st...  123.88KB
2  hdfs://localhost:20500/test-warehouse/tpcds.st...  139.28KB
3  hdfs://localhost:20500/test-warehouse/tpcds.st...  139.60KB
4  hdfs://localhost:20500/test-warehouse/tpcds.st...   62.84KB

                 partition
0  ss_sold_date_sk=2451803
1  ss_sold_date_sk=2451819
2  ss_sold_date_sk=2451772
3  ss_sold_date_sk=2451789
4  ss_sold_date_sk=2451741

Modifying table metadata

For unpartitioned tables, you can use the alter method to change its location, file format, and other properties. For partitioned tables, to change partition-specific metadata use alter_partition.

ImpalaTable.alter([location, format, ...])

Change setting and parameters of the table.

ImpalaTable.alter_partition(spec[, ...])

Change setting and parameters of an existing partition

For example, if you wanted to “point” an existing table at a directory of CSV files, you could run the following command:

from getpass import getuser

csv_props = {
    'serialization.format': ',',
    'field.delim': ',',
}
data_dir = '/home/{}/my-csv-files'.format(getuser())

table.alter(location=data_dir, format='text', serde_properties=csv_props)

If the table is partitioned, you can modify only the properties of a particular partition:

table.alter_partition(
    {'year': 2007, 'month': 5},
    location=data_dir,
    format='text',
    serde_properties=csv_props
)

Table statistics

Computing table and partition statistics

ImpalaTable.compute_stats([incremental])

Invoke Impala COMPUTE STATS command to compute column, table, and partition statistics.

Impala-backed physical tables have a method compute_stats that computes table, column, and partition-level statistics to assist with query planning and optimization. It is standard practice to invoke this after creating a table or loading new data:

table.compute_stats()

If you are using a recent version of Impala, you can also access the COMPUTE INCREMENTAL STATS DDL command:

table.compute_stats(incremental=True)

Seeing table and column statistics

ImpalaTable.column_stats()

Return results of SHOW COLUMN STATS as a pandas DataFrame

ImpalaTable.stats()

Return results of SHOW TABLE STATS as a DataFrame.

The compute_stats and stats functions return the results of SHOW COLUMN STATS and SHOW TABLE STATS, respectively, and their output will depend, of course, on the last COMPUTE STATS call.

>>> ss = c.table('tpcds_parquet.store_sales')
>>> ss.compute_stats(incremental=True)
>>> stats = ss.stats()
>>> stats[:5]
  ss_sold_date_sk  #Rows  #Files     Size Bytes Cached Cache Replication  \
0         2450829   1071       1  78.34KB   NOT CACHED        NOT CACHED
1         2450846    839       1  61.83KB   NOT CACHED        NOT CACHED
2         2450860    747       1  54.86KB   NOT CACHED        NOT CACHED
3         2450874    922       1  66.74KB   NOT CACHED        NOT CACHED
4         2450888    856       1  63.33KB   NOT CACHED        NOT CACHED

    Format Incremental stats  \
0  PARQUET              true
1  PARQUET              true
2  PARQUET              true
3  PARQUET              true
4  PARQUET              true

                                            Location
0  hdfs://localhost:20500/test-warehouse/tpcds.st...
1  hdfs://localhost:20500/test-warehouse/tpcds.st...
2  hdfs://localhost:20500/test-warehouse/tpcds.st...
3  hdfs://localhost:20500/test-warehouse/tpcds.st...
4  hdfs://localhost:20500/test-warehouse/tpcds.st...

>>> cstats = ss.column_stats()
>>> cstats
                   Column          Type  #Distinct Values  #Nulls  Max Size  Avg Size
0         ss_sold_time_sk        BIGINT             13879      -1       NaN         8
1              ss_item_sk        BIGINT             17925      -1       NaN         8
2          ss_customer_sk        BIGINT             15207      -1       NaN         8
3             ss_cdemo_sk        BIGINT             16968      -1       NaN         8
4             ss_hdemo_sk        BIGINT              6220      -1       NaN         8
5              ss_addr_sk        BIGINT             14077      -1       NaN         8
6             ss_store_sk        BIGINT                 6      -1       NaN         8
7             ss_promo_sk        BIGINT               298      -1       NaN         8
8        ss_ticket_number           INT             15006      -1       NaN         4
9             ss_quantity           INT                99      -1       NaN         4
10      ss_wholesale_cost  DECIMAL(7,2)             10196      -1       NaN         4
11          ss_list_price  DECIMAL(7,2)             19393      -1       NaN         4
12         ss_sales_price  DECIMAL(7,2)             15594      -1       NaN         4
13    ss_ext_discount_amt  DECIMAL(7,2)             29772      -1       NaN         4
14     ss_ext_sales_price  DECIMAL(7,2)            102758      -1       NaN         4
15  ss_ext_wholesale_cost  DECIMAL(7,2)            125448      -1       NaN         4
16      ss_ext_list_price  DECIMAL(7,2)            141419      -1       NaN         4
17             ss_ext_tax  DECIMAL(7,2)             33837      -1       NaN         4
18          ss_coupon_amt  DECIMAL(7,2)             29772      -1       NaN         4
19            ss_net_paid  DECIMAL(7,2)            109981      -1       NaN         4
20    ss_net_paid_inc_tax  DECIMAL(7,2)            132286      -1       NaN         4
21          ss_net_profit  DECIMAL(7,2)            122436      -1       NaN         4
22        ss_sold_date_sk        BIGINT               120       0       NaN         8

REFRESH and INVALIDATE METADATA

These DDL commands are available as table-level and client-level methods:

ImpalaClient.invalidate_metadata([name, ...])

Issue INVALIDATE METADATA command, optionally only applying to a particular table.

ImpalaTable.invalidate_metadata()

ImpalaTable.refresh()

You can invalidate the cached metadata for a single table or for all tables using invalidate_metadata, and similarly invoke REFRESH db_name.table_name using the refresh method.

client.invalidate_metadata()

table = db.table(table_name)
table.invalidate_metadata()

table.refresh()

These methods are often used in conjunction with the LOAD DATA commands and COMPUTE STATS. See the Impala documentation for full details.

Issuing LOAD DATA commands

The LOAD DATA DDL physically moves a single data file or a directory of files into the correct location for a table or table partition. It is especially useful for partitioned tables as you do not have to construct the directory path for a partition by hand, so simpler and less error-prone than manually moving files with low level HDFS commands. It also deals with file name conflicts so data is not lost in such cases.

ImpalaClient.load_data(table_name, path[, ...])

Wraps the LOAD DATA DDL statement.

ImpalaTable.load_data(path[, overwrite, ...])

Wraps the LOAD DATA DDL statement.

To use these methods, pass the path of a single file or a directory of files you want to load. Afterward, you may want to update the table statistics (see Impala documentation):

table.load_data(path)
table.refresh()

Like the other methods with support for partitioned tables, you can load into a particular partition with the partition keyword argument:

part = [2007, 5]
table.load_data(path, partition=part)

Parquet and other session options

Ibis gives you access to Impala session-level variables that affect query execution:

ImpalaClient.disable_codegen([disabled])

Turn off or on LLVM codegen in Impala query execution

ImpalaClient.get_options()

Return current query options for the Impala session

ImpalaClient.set_options(options)

ImpalaClient.set_compression_codec(codec)

Parameters

For example:

>>> client.get_options()
{'ABORT_ON_ERROR': '0',
 'APPX_COUNT_DISTINCT': '0',
 'BUFFER_POOL_LIMIT': '',
 'COMPRESSION_CODEC': '',
 'COMPUTE_STATS_MIN_SAMPLE_SIZE': '1073741824',
 'DEFAULT_JOIN_DISTRIBUTION_MODE': '0',
 'DEFAULT_SPILLABLE_BUFFER_SIZE': '2097152',
 'DISABLE_CODEGEN': '0',
 'DISABLE_CODEGEN_ROWS_THRESHOLD': '50000',
 'DISABLE_ROW_RUNTIME_FILTERING': '0',
 'DISABLE_STREAMING_PREAGGREGATIONS': '0',
 'DISABLE_UNSAFE_SPILLS': '0',
 'ENABLE_EXPR_REWRITES': '1',
 'EXEC_SINGLE_NODE_ROWS_THRESHOLD': '100',
 'EXEC_TIME_LIMIT_S': '0',
 'EXPLAIN_LEVEL': '1',
 'HBASE_CACHE_BLOCKS': '0',
 'HBASE_CACHING': '0',
 'IDLE_SESSION_TIMEOUT': '0',
 'MAX_ERRORS': '100',
 'MAX_NUM_RUNTIME_FILTERS': '10',
 'MAX_ROW_SIZE': '524288',
 'MEM_LIMIT': '0',
 'MIN_SPILLABLE_BUFFER_SIZE': '65536',
 'MT_DOP': '',
 'NUM_SCANNER_THREADS': '0',
 'OPTIMIZE_PARTITION_KEY_SCANS': '0',
 'PARQUET_ANNOTATE_STRINGS_UTF8': '0',
 'PARQUET_ARRAY_RESOLUTION': '2',
 'PARQUET_DICTIONARY_FILTERING': '1',
 'PARQUET_FALLBACK_SCHEMA_RESOLUTION': '0',
 'PARQUET_FILE_SIZE': '0',
 'PARQUET_READ_STATISTICS': '1',
 'PREFETCH_MODE': '1',
 'QUERY_TIMEOUT_S': '0',
 'REPLICA_PREFERENCE': '0',
 'REQUEST_POOL': '',
 'RUNTIME_BLOOM_FILTER_SIZE': '1048576',
 'RUNTIME_FILTER_MAX_SIZE': '16777216',
 'RUNTIME_FILTER_MIN_SIZE': '1048576',
 'RUNTIME_FILTER_MODE': '2',
 'RUNTIME_FILTER_WAIT_TIME_MS': '0',
 'S3_SKIP_INSERT_STAGING': '1',
 'SCHEDULE_RANDOM_REPLICA': '0',
 'SCRATCH_LIMIT': '-1',
 'SEQ_COMPRESSION_MODE': '',
 'SYNC_DDL': '0'}

To enable Snappy compression for Parquet files, you could do either of:

>>> client.set_options({'COMPRESSION_CODEC': 'snappy'})
>>> client.set_compression_codec('snappy')

>>> client.get_options()['COMPRESSION_CODEC']
'SNAPPY'

Ingesting data from pandas

Overall interoperability between the Hadoop / Spark ecosystems and pandas / the PyData stack is poor, but it will improve in time (this is a major part of the Ibis roadmap).

Ibis’s Impala tools currently interoperate with pandas in these ways:

  • Ibis expressions return pandas objects (i.e. DataFrame or Series) for non-scalar expressions when calling their execute method

  • The create_table and insert methods can accept pandas objects. This includes inserting into partitioned tables. It currently uses CSV as the ingest route.

For example:

>>> import pandas as pd

>>> data = pd.DataFrame({'foo': [1, 2, 3, 4], 'bar': ['a', 'b', 'c', 'd']})

>>> db.create_table('pandas_table', data)
>>> t = db.pandas_table
>>> t.execute()
  bar  foo
0   a    1
1   b    2
2   c    3
3   d    4

>>> t.drop()

>>> db.create_table('empty_for_insert', schema=t.schema())

>>> to_insert = db.empty_for_insert
>>> to_insert.insert(data)
>>> to_insert.execute()
  bar  foo
0   a    1
1   b    2
2   c    3
3   d    4

>>> to_insert.drop()
>>> import pandas as pd

>>> data = pd.DataFrame({'foo': [1, 2, 3, 4], 'bar': ['a', 'b', 'c', 'd']})

>>> db.create_table('pandas_table', data)
>>> t = db.pandas_table
>>> t.execute()
   foo bar
0    1   a
1    2   b
2    3   c
3    4   d

>>> t.drop()
>>> db.create_table('empty_for_insert', schema=t.schema())
>>> to_insert = db.empty_for_insert
>>> to_insert.insert(data)
>>> to_insert.execute()
   foo bar
0    1   a
1    2   b
2    3   c
3    4   d

>>> to_insert.drop()

Uploading / downloading data from HDFS

If you’ve set up an HDFS connection, you can use the Ibis HDFS interface to look through your data and read and write files to and from HDFS:

>>> hdfs = con.hdfs
>>> hdfs.ls('/__ibis/ibis-testing-data')
['README.md',
 'avro',
 'awards_players.csv',
 'batting.csv',
 'csv',
 'diamonds.csv',
 'functional_alltypes.csv',
 'functional_alltypes.parquet',
 'geo.csv',
 'ibis_testing.db',
 'parquet',
 'struct_table.avro',
 'udf']
>>> hdfs.ls('/__ibis/ibis-testing-data/parquet')
['functional_alltypes',
 'tpch_customer',
 'tpch_lineitem',
 'tpch_nation',
 'tpch_orders',
 'tpch_part',
 'tpch_partsupp',
 'tpch_region',
 'tpch_supplier']

Suppose we wanted to download /__ibis/ibis-testing-data/parquet/functional_alltypes, which is a directory. We need only do:

$ rm -rf parquet_dir/
>>> hdfs.get('/__ibis/ibis-testing-data/parquet/functional_alltypes',
...          'parquet_dir')
'/ibis/docs/source/tutorial/parquet_dir'

Now we have that directory locally:

$ ls parquet_dir/
9a41de519352ab07-4e76bc4d9fb5a789_1624886651_data.0.parq
9a41de519352ab07-4e76bc4d9fb5a78a_778826485_data.0.parq
9a41de519352ab07-4e76bc4d9fb5a78b_1277612014_data.0.parq

Files and directories can be written to HDFS just as easily using put:

>>> path = '/__ibis/dir-write-example'
>>> if hdfs.exists(path):
...    hdfs.rmdir(path)
>>> hdfs.put(path, 'parquet_dir', verbose=True)
'/__ibis/dir-write-example'
>>> hdfs.ls('/__ibis/dir-write-example')
['9a41de519352ab07-4e76bc4d9fb5a789_1624886651_data.0.parq',
 '9a41de519352ab07-4e76bc4d9fb5a78a_778826485_data.0.parq',
 '9a41de519352ab07-4e76bc4d9fb5a78b_1277612014_data.0.parq']

Delete files with rm or directories with rmdir:

>>> hdfs.rmdir('/__ibis/dir-write-example')
rm -rf parquet_dir/

Queries on Parquet, Avro, and Delimited files in HDFS

Ibis can easily create temporary or persistent Impala tables that reference data in the following formats:

  • Parquet (parquet_file)

  • Avro (avro_file)

  • Delimited text formats (CSV, TSV, etc.) (delimited_file)

Parquet is the easiest because the schema can be read from the data files:

>>> path = '/__ibis/ibis-testing-data/parquet/tpch_lineitem'
>>> lineitem = con.parquet_file(path)
>>> lineitem.limit(2)
   l_orderkey  l_partkey  l_suppkey  l_linenumber l_quantity l_extendedprice  \
0           1     155190       7706             1      17.00        21168.23
1           1      67310       7311             2      36.00        45983.16

  l_discount l_tax l_returnflag l_linestatus  l_shipdate l_commitdate  \
0       0.04  0.02            N            O  1996-03-13   1996-02-12
1       0.09  0.06            N            O  1996-04-12   1996-02-28

  l_receiptdate     l_shipinstruct l_shipmode  \
0    1996-03-22  DELIVER IN PERSON      TRUCK
1    1996-04-20   TAKE BACK RETURN       MAIL

                            l_comment
0             egular courts above the
1  ly final dependencies: slyly bold
>>> lineitem.l_extendedprice.sum()
Decimal('229577310901.20')

If you want to query a Parquet file and also create a table in Impala that remains after your session, you can pass more information to parquet_file:

>>> table = con.parquet_file(path, name='my_parquet_table',
...                          database='ibis_testing',
...                          persist=True)
>>> table.l_extendedprice.sum()
Decimal('229577310901.20')
>>> con.table('my_parquet_table').l_extendedprice.sum()
Decimal('229577310901.20')
>>> con.drop_table('my_parquet_table')

To query delimited files, you need to write down an Ibis schema. At some point we’d like to build some helper tools that will infer the schema for you, all in good time.

There’s some CSV files in the test folder, so let’s use those:

>>> hdfs.get('/__ibis/ibis-testing-data/csv', 'csv-files')
'/ibis/docs/source/tutorial/csv-files'
$ cat csv-files/0.csv
63IEbRheTh,0.679388707915,6
mG4hlqnjeG,2.80710565922,15
JTPdX9SZH5,-0.155126406372,55
2jcl6FypOl,1.03787834032,21
k3TbJLaadQ,-1.40190801103,23
rP5J4xvinM,-0.442092712869,22
WniUylixYt,-0.863748033806,27
znsDuKOB1n,-0.566029637098,47
4SRP9jlo1M,0.331460412318,88
KsfjPyDf5e,-0.578930506363,70
$ rm -rf csv-files/

The schema here is pretty simple (see ibis.schema for more):

>>> schema = ibis.schema([('foo', 'string'),
...                       ('bar', 'double'),
...                       ('baz', 'int32')])

>>> table = con.delimited_file('/__ibis/ibis-testing-data/csv',
...                            schema)
>>> table.limit(10)
          foo       bar  baz
0  63IEbRheTh  0.679389    6
1  mG4hlqnjeG  2.807106   15
2  JTPdX9SZH5 -0.155126   55
3  2jcl6FypOl  1.037878   21
4  k3TbJLaadQ -1.401908   23
5  rP5J4xvinM -0.442093   22
6  WniUylixYt -0.863748   27
7  znsDuKOB1n -0.566030   47
8  4SRP9jlo1M  0.331460   88
9  KsfjPyDf5e -0.578931   70
>>> table.bar.summary()
   count  nulls       min       max       sum    mean  approx_nunique
0    100      0 -1.401908  2.807106  8.479978  0.0848              10

For functions like parquet_file and delimited_file, an HDFS directory must be passed (we’ll add support for S3 and other filesystems later) and the directory must contain files all having the same schema.

If you have Avro data, you can query it too if you have the full avro schema:

>>> avro_schema = {
...     "fields": [
...         {"type": ["int", "null"], "name": "R_REGIONKEY"},
...         {"type": ["string", "null"], "name": "R_NAME"},
...         {"type": ["string", "null"], "name": "R_COMMENT"}],
...     "type": "record",
...     "name": "a"
... }

>>> path = '/__ibis/ibis-testing-data/avro/tpch.region'

>>> hdfs.mkdir(path)
>>> table = con.avro_file(path, avro_schema)
>>> table
Empty DataFrame
Columns: [r_regionkey, r_name, r_comment]
Index: []

Other helper functions for interacting with the database

We’re adding a growing list of useful utility functions for interacting with an Impala cluster on the client object. The idea is that you should be able to do any database-admin-type work with Ibis and not have to switch over to the Impala SQL shell. Any ways we can make this more pleasant, please let us know.

Here’s some of the features, which we’ll give examples for:

  • Listing and searching for available databases and tables

  • Creating and dropping databases

  • Getting table schemas

>>> con.list_databases(like='ibis*')
['ibis_testing', 'ibis_testing_tmp_db']
>>> con.list_tables(database='ibis_testing', like='tpch*')
['tpch_customer',
 'tpch_lineitem',
 'tpch_nation',
 'tpch_orders',
 'tpch_part',
 'tpch_partsupp',
 'tpch_region',
 'tpch_region_avro',
 'tpch_supplier']
>>> schema = con.get_schema('functional_alltypes')
>>> schema
ibis.Schema {
  id               int32
  bool_col         boolean
  tinyint_col      int8
  smallint_col     int16
  int_col          int32
  bigint_col       int64
  float_col        float32
  double_col       float64
  date_string_col  string
  string_col       string
  timestamp_col    timestamp
  year             int32
  month            int32
}

Databases can be created, too, and you can set the storage path in HDFS you want for the data files

>>> db = 'ibis_testing2'
>>> con.create_database(db, path='/__ibis/my-test-database', force=True)

>>> # you may or may not have to give the impala user write and execute permissions to '/__ibis/my-test-database'
>>> hdfs.chmod('/__ibis/my-test-database', '777')
>>> con.create_table('example_table', con.table('functional_alltypes'),
...                  database=db, force=True)

Hopefully, there will be data files in the indicated spot in HDFS:

>>> hdfs.ls('/__ibis/my-test-database')
['example_table']

To drop a database, including all tables in it, you can use drop_database with force=True:

>>> con.drop_database(db, force=True)

Faster queries on small data in Impala

Since Impala internally uses LLVM to compile parts of queries (aka “codegen”) to make them faster on large data sets there is a certain amount of overhead with running many kinds of queries, even on small datasets. You can disable LLVM code generation when using Ibis, which may significantly speed up queries on smaller datasets:

>>> from numpy.random import rand
>>> con.disable_codegen()
>>> t = con.table('ibis_testing.functional_alltypes')
$ time python -c "(t.double_col + rand()).sum().execute()"
27.7 ms ± 996 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)
# Turn codegen back on
con.disable_codegen(False)
$ time python -c "(t.double_col + rand()).sum().execute()"
27 ms ± 1.62 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)

It’s important to remember that codegen is a fixed overhead and will significantly speed up queries on big data

User Defined functions (UDF)

Impala currently supports user-defined scalar functions (known henceforth as UDFs) and aggregate functions (respectively UDAs) via a C++ extension API.

Initial support for using C++ UDFs in Ibis came in version 0.4.0.

Using scalar functions (UDFs)

Let’s take an example to illustrate how to make a C++ UDF available to Ibis. Here is a function that computes an approximate equality between floating point values:

#include "impala_udf/udf.h"

#include <cctype>
#include <cmath>

BooleanVal FuzzyEquals(FunctionContext* ctx, const DoubleVal& x, const DoubleVal& y) {
  const double EPSILON = 0.000001f;
  if (x.is_null || y.is_null) return BooleanVal::null();
  double delta = fabs(x.val - y.val);
  return BooleanVal(delta < EPSILON);
}

You can compile this to either a shared library (a .so file) or to LLVM bitcode with clang (a .ll file). Skipping that step for now (will add some more detailed instructions here later, promise).

To make this function callable, we use ibis.impala.wrap_udf:

library = '/ibis/udfs/udftest.ll'
inputs = ['double', 'double']
output = 'boolean'
symbol = 'FuzzyEquals'
udf_db = 'ibis_testing'
udf_name = 'fuzzy_equals'

fuzzy_equals = ibis.impala.wrap_udf(
    library, inputs, output, symbol, name=udf_name
)

In typical workflows, you will set up a UDF in Impala once then use it thenceforth. So the first time you do this, you need to create the UDF in Impala:

client.create_function(fuzzy_equals, database=udf_db)

Now, we must register this function as a new Impala operation in Ibis. This must take place each time you load your Ibis session.

func.register(fuzzy_equals.name, udf_db)

The object fuzzy_equals is callable and works with Ibis expressions:

>>> db = c.database('ibis_testing')

>>> t = db.functional_alltypes

>>> expr = fuzzy_equals(t.float_col, t.double_col / 10)

>>> expr.execute()[:10]
0     True
1    False
2    False
3    False
4    False
5    False
6    False
7    False
8    False
9    False
Name: tmp, dtype: bool

Note that the call to register on the UDF object must happen each time you use Ibis. If you have a lot of UDFs, I suggest you create a file with all of your wrapper declarations and user APIs that you load with your Ibis session to plug in all your own functions.

Adding documentation to new functions

fuzzy_equal.__doc__ = """\
Approximate equals UDF

Parameters
----------
left : numeric
right : numeric

Returns
-------
bool

"""