Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Minor: Document parquet_metadata function #8852

Merged
merged 2 commits into from
Jan 14, 2024
Merged
Changes from 1 commit
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
69 changes: 68 additions & 1 deletion docs/source/user-guide/cli.md
Original file line number Diff line number Diff line change
Expand Up @@ -191,7 +191,7 @@ DataFusion CLI v16.0.0
2 rows in set. Query took 0.007 seconds.
```

## Creating external tables
## Creating External Tables
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

drive by cleanup -- the other headings are capitalized so it seemed strange that this one was not


It is also possible to create a table backed by files by explicitly
via `CREATE EXTERNAL TABLE` as shown below. Filemask wildcards supported
Expand Down Expand Up @@ -425,6 +425,13 @@ Available commands inside DataFusion CLI are:
> \h function
```

## Supported SQL

In addition to the normal [SQL supported in DataFusion], `datatfusion-cli` also
alamb marked this conversation as resolved.
Show resolved Hide resolved
supports additional statements and commands:

[sql supported in datafusion]: sql/index.rst

- Show configuration options

`SHOW ALL [VERBOSE]`
Expand Down Expand Up @@ -467,6 +474,66 @@ Available commands inside DataFusion CLI are:
> SET datafusion.execution.batch_size to 1024;
```

- `parquet_metadata` table function
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Here is what this looks like rendered:
Screenshot 2024-01-13 at 7 58 34 AM


The `parquet_metadata` table function can be used to inspect detailed metadata
about a parquet file such as statistics, sizes, and other information. This can
be helpful to understand how parquet files are structured.

For example, to see information about the `"WatchID"` column in the
`hits.parquet` file, you can use:

```sql
SELECT path_in_schema, row_group_id, row_group_num_rows, stats_min, stats_max, total_compressed_size
FROM parquet_metadata('hits.parquet')
WHERE path_in_schema = '"WatchID"'
LIMIT 3;

+----------------+--------------+--------------------+---------------------+---------------------+-----------------------+
| path_in_schema | row_group_id | row_group_num_rows | stats_min | stats_max | total_compressed_size |
+----------------+--------------+--------------------+---------------------+---------------------+-----------------------+
| "WatchID" | 0 | 450560 | 4611687214012840539 | 9223369186199968220 | 3883759 |
| "WatchID" | 1 | 612174 | 4611689135232456464 | 9223371478009085789 | 5176803 |
| "WatchID" | 2 | 344064 | 4611692774829951781 | 9223363791697310021 | 3031680 |
+----------------+--------------+--------------------+---------------------+---------------------+-----------------------+
3 rows in set. Query took 0.053 seconds.
```

The returned table has the following columns for each row for each column chunk
in the file. Please refer to the [Parquet Documentation] for more information.

[parquet documentation]: https://parquet.apache.org/

| column_name | data_type | Description |
| ----------------------- | --------- | --------------------------------------------------------------------------------------------------- |
| filename | Utf8 | Name of the file |
| row_group_id | Int64 | Row group index the column chunk belongs to |
| row_group_num_rows | Int64 | Count of rows stored in the row group |
| row_group_num_columns | Int64 | Total number of columns in the row group (same for all row groups) |
| row_group_bytes | Int64 | Number of bytes used to store the row group (not including metadata) |
| column_id | Int64 | ID of the column |
| file_offset | Int64 | Offset within the file that this column chunk's data begins |
| num_values | Int64 | Total number of values in this column chunk |
| path_in_schema | Utf8 | "Path" (column name) of the column chunk in the schema |
| type | Utf8 | Parquet data type of the column chunk |
| stats_min | Utf8 | The minimum value for this column chunk, if stored in the statistics, cast to a string |
| stats_max | Utf8 | The minimum value for this column chunk, if stored in the statistics, cast to a string |
alamb marked this conversation as resolved.
Show resolved Hide resolved
| stats_null_count | Int64 | Number of null values in this column chunk, if stored in the statistics |
| stats_distinct_count | Int64 | Number of distinct values in this column chunk, if stored in the statistics |
| stats_min_value | Utf8 | Same as `stats_min` |
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

wondering if this duplicated fields needed?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't honestly know why the seemingly duplicated columns are present. It was done initially to mirror duckdb which has them. Maybe we should investigate the reason why 🤔

D create table foo as select * from parquet_metadata('./benchmarks/data/hits.parquet');
D describe table foo;
┌─────────────────────────┬─────────────┬─────────┬─────────┬─────────┬─────────┐
│       column_name       │ column_type │  null   │   key   │ default │  extra  │
│         varcharvarcharvarcharvarcharvarcharvarchar │
├─────────────────────────┼─────────────┼─────────┼─────────┼─────────┼─────────┤
│ file_name               │ VARCHAR     │ YES     │         │         │         │
│ row_group_id            │ BIGINT      │ YES     │         │         │         │
│ row_group_num_rows      │ BIGINT      │ YES     │         │         │         │
│ row_group_num_columns   │ BIGINT      │ YES     │         │         │         │
│ row_group_bytes         │ BIGINT      │ YES     │         │         │         │
│ column_id               │ BIGINT      │ YES     │         │         │         │
│ file_offset             │ BIGINT      │ YES     │         │         │         │
│ num_values              │ BIGINT      │ YES     │         │         │         │
│ path_in_schema          │ VARCHAR     │ YES     │         │         │         │
│ type                    │ VARCHAR     │ YES     │         │         │         │
│ stats_min               │ VARCHAR     │ YES     │         │         │         │
│ stats_max               │ VARCHAR     │ YES     │         │         │         │
│ stats_null_count        │ BIGINT      │ YES     │         │         │         │
│ stats_distinct_count    │ BIGINT      │ YES     │         │         │         │
│ stats_min_value         │ VARCHAR     │ YES     │         │         │         │
│ stats_max_value         │ VARCHAR     │ YES     │         │         │         │
│ compression             │ VARCHAR     │ YES     │         │         │         │
│ encodings               │ VARCHAR     │ YES     │         │         │         │
│ index_page_offset       │ BIGINT      │ YES     │         │         │         │
│ dictionary_page_offset  │ BIGINT      │ YES     │         │         │         │
│ data_page_offset        │ BIGINT      │ YES     │         │         │         │
│ total_compressed_size   │ BIGINT      │ YES     │         │         │         │
│ total_uncompressed_size │ BIGINT      │ YES     │         │         │         │
├─────────────────────────┴─────────────┴─────────┴─────────┴─────────┴─────────┤
│ 23 rows                                                             6 columns │
└───────────────────────────────────────────────────────────────────────────────┘

| stats_max_value | Utf8 | Same as `stats_max` |
| compression | Utf8 | Block level compression (e.g. `SNAPPY`) used for this column chunk |
| encodings | Utf8 | All block level encodings (e.g. `[PLAIN_DICTIONARY, PLAIN, RLE]`) used for this column chunk |
| index_page_offset | Int64 | Offset in the file of the [`page index`], if any |
| dictionary_page_offset | Int64 | Offset in the file of the dictionary page, if any |
| data_page_offset | Int64 | Offset in the file of the first data page, if any |
| total_compressed_size | Int64 | Number of bytes the column chunk's data after encoding and compression (what is stored in the file) |
| total_uncompressed_size | Int64 | Number of bytes the column chunk's data after encoding |

+-------------------------+-----------+-------------+

[`page index`]: https://github.com/apache/parquet-format/blob/master/PageIndex.md

## Changing Configuration Options

All available configuration options can be seen using `SHOW ALL` as described above.
Expand Down