feat: DataFrame optimalization

CHANGELOG.md

@@ -2,6 +2,7 @@
 
 ### Features
 1. [#79](https://github.com/influxdata/influxdb-client-python/issues/79): Added support for writing Pandas DataFrame
+2. [#92](https://github.com/influxdata/influxdb-client-python/issues/92): Optimize serializing Pandas DataFrame for writing
 
 ### Bug Fixes
 1. [#85](https://github.com/influxdata/influxdb-client-python/issues/85): Fixed a possibility to generate empty write batch

extra-requirements.txt

@@ -1 +1,2 @@
 pandas>=0.25.3
+numpy

influxdb_client/client/write_api.py

@@ -1,8 +1,11 @@
 # coding: utf-8
 import logging
 import os
+import re
 from datetime import timedelta
 from enum import Enum
+from functools import reduce
+from itertools import chain
 from random import random
 from time import sleep
 from typing import Union, List
@@ -14,7 +17,7 @@
 
 from influxdb_client import WritePrecision, WriteService
 from influxdb_client.client.abstract_client import AbstractClient
-from influxdb_client.client.write.point import Point, DEFAULT_WRITE_PRECISION
+from influxdb_client.client.write.point import Point, DEFAULT_WRITE_PRECISION, _ESCAPE_KEY
 from influxdb_client.rest import ApiException
 
 logger = logging.getLogger(__name__)
@@ -253,10 +256,8 @@ def _serialize(self, record, write_precision, **kwargs) -> bytes:
             _result = self._serialize(Point.from_dict(record, write_precision=write_precision),
                                       write_precision, **kwargs)
         elif 'DataFrame' in type(record).__name__:
-            _result = self._serialize(self._data_frame_to_list_of_points(record,
-                                                                         precision=write_precision, **kwargs),
-                                      write_precision,
-                                      **kwargs)
+            _data = self._data_frame_to_list_of_points(record, precision=write_precision, **kwargs)
+            _result = self._serialize(_data, write_precision, **kwargs)
 
         elif isinstance(record, list):
             _result = b'\n'.join([self._serialize(item, write_precision,
@@ -297,8 +298,15 @@ def _write_batching(self, bucket, org, data,
 
         return None
 
+    def _itertuples(self, data_frame):
+        """Custom implementation of ``DataFrame.itertuples`` that
+        returns plain tuples instead of namedtuples. About 50% faster.
+        """
+        cols = [data_frame.iloc[:, k] for k in range(len(data_frame.columns))]
+        return zip(data_frame.index, *cols)
+
     def _data_frame_to_list_of_points(self, data_frame, precision, **kwargs):
-        from ..extras import pd
+        from ..extras import pd, np
         if not isinstance(data_frame, pd.DataFrame):
             raise TypeError('Must be DataFrame, but type was: {0}.'
                             .format(type(data_frame)))
@@ -314,28 +322,35 @@ def _data_frame_to_list_of_points(self, data_frame, precision, **kwargs):
         if data_frame.index.tzinfo is None:
             data_frame.index = data_frame.index.tz_localize('UTC')
 
-        data = []
+        measurement_name = kwargs.get('data_frame_measurement_name')
+        data_frame_tag_columns = kwargs.get('data_frame_tag_columns')
+        data_frame_tag_columns = set(data_frame_tag_columns or [])
 
-        for c, (row) in enumerate(data_frame.values):
-            point = Point(measurement_name=kwargs.get('data_frame_measurement_name'))
+        tags = []
+        fields = []
 
-            for count, (value) in enumerate(row):
-                column = data_frame.columns[count]
-                data_frame_tag_columns = kwargs.get('data_frame_tag_columns')
-                if data_frame_tag_columns and column in data_frame_tag_columns:
-                    point.tag(column, value)
-                else:
-                    point.field(column, value)
+        if self._point_settings.defaultTags:
+            for key, value in self._point_settings.defaultTags.items():
+                data_frame[key] = value
+                data_frame_tag_columns.add(key)
 
-            point.time(data_frame.index[c], precision)
+        for index, (key, value) in enumerate(data_frame.dtypes.items()):
+            key = str(key).translate(_ESCAPE_KEY)
 
-            if self._point_settings.defaultTags:
-                for key, val in self._point_settings.defaultTags.items():
-                    point.tag(key, val)
+            if key in data_frame_tag_columns:
+                tags.append(f"{key}={{p[{index + 1}].translate(_ESCAPE_KEY)}}")
+            elif issubclass(value.type, np.integer):
+                fields.append(f"{key}={{p[{index + 1}]}}i")
+            elif issubclass(value.type, (np.float, np.bool_)):
+                fields.append(f"{key}={{p[{index + 1}]}}")
+            else:
+                fields.append(f"{key}=\"{{p[{index + 1}].translate(_ESCAPE_KEY)}}\"")
 
-            data.append(point)
+        fmt = (f'{measurement_name}', f'{"," if tags else ""}', ','.join(tags),
+               ' ', ','.join(fields), ' {p[0].value}')
+        f = eval("lambda p: f'{}'".format(''.join(fmt)))
 
-        return data
+        return list(map(f, self._itertuples(data_frame)))
 
     def _http(self, batch_item: _BatchItem):
 

influxdb_client/extras.py

@@ -3,4 +3,9 @@
 except ModuleNotFoundError as err:
     raise ImportError(f"`query_data_frame` requires Pandas which couldn't be imported due: {err}")
 
-__all__ = ['pd']
+try:
+    import numpy as np
+except ModuleNotFoundError as err:
+    raise ImportError(f"`data_frame` requires numpy which couldn't be imported due: {err}")
+
+__all__ = ['pd', 'np']

tests/test_WriteApi.py

@@ -231,9 +231,9 @@ def test_write_data_frame(self):
         bucket = self.create_test_bucket()
 
         now = pd.Timestamp('1970-01-01 00:00+00:00')
-        data_frame = pd.DataFrame(data=[["coyote_creek", 1.0], ["coyote_creek", 2.0]],
+        data_frame = pd.DataFrame(data=[["coyote_creek", 1], ["coyote_creek", 2]],
                                   index=[now + timedelta(hours=1), now + timedelta(hours=2)],
-                                  columns=["location", "water_level"])
+                                  columns=["location", "water water_level"])
 
         self.write_client.write(bucket.name, record=data_frame, data_frame_measurement_name='h2o_feet',
                                 data_frame_tag_columns=['location'])
@@ -247,14 +247,14 @@ def test_write_data_frame(self):
         self.assertEqual(result[0].records[0].get_measurement(), "h2o_feet")
         self.assertEqual(result[0].records[0].get_value(), 1.0)
         self.assertEqual(result[0].records[0].values.get("location"), "coyote_creek")
-        self.assertEqual(result[0].records[0].get_field(), "water_level")
+        self.assertEqual(result[0].records[0].get_field(), "water water_level")
         self.assertEqual(result[0].records[0].get_time(),
                          datetime.datetime(1970, 1, 1, 1, 0, tzinfo=datetime.timezone.utc))
 
         self.assertEqual(result[0].records[1].get_measurement(), "h2o_feet")
         self.assertEqual(result[0].records[1].get_value(), 2.0)
         self.assertEqual(result[0].records[1].values.get("location"), "coyote_creek")
-        self.assertEqual(result[0].records[1].get_field(), "water_level")
+        self.assertEqual(result[0].records[1].get_field(), "water water_level")
         self.assertEqual(result[0].records[1].get_time(),
                          datetime.datetime(1970, 1, 1, 2, 0, tzinfo=datetime.timezone.utc))
 

tests/test_WriteApiDataFrame.py

@@ -0,0 +1,53 @@
+import csv
+import os
+import time
+import unittest
+
+from influxdb_client import InfluxDBClient, WriteOptions, WriteApi, Point
+from tests.base_test import BaseTest
+
+
+class DataFrameWriteTest(BaseTest):
+
+    def setUp(self) -> None:
+        self.influxDb_client = InfluxDBClient(url="http://localhost:9999", token="my-token")
+
+        self.write_options = WriteOptions(batch_size=10_000, flush_interval=5_000, retry_interval=3_000)
+        self._write_client = WriteApi(influxdb_client=self.influxDb_client, write_options=self.write_options)
+
+    def tearDown(self) -> None:
+        self._write_client.__del__()
+
+    @unittest.skip('Test big file')
+    def test_write_data_frame(self):
+        import random
+        from influxdb_client.extras import pd
+
+        if not os.path.isfile("data_frame_file.csv"):
+            with open('data_frame_file.csv', mode='w+') as csv_file:
+                _writer = csv.writer(csv_file, delimiter=',', quotechar='"', quoting=csv.QUOTE_MINIMAL)
+                _writer.writerow(['time', 'col1', 'col2', 'col3', 'col4', 'col5', 'col6', 'col7', 'col8'])
+
+                for i in range(1, 1500000):
+                    choice = ['test_a', 'test_b', 'test_c']
+                    _writer.writerow([i, random.choice(choice), 'test', random.random(), random.random(),
+                                      random.random(), random.random(), random.random(), random.random()])
+
+            csv_file.close()
+
+        with open('data_frame_file.csv', mode='rb') as csv_file:
+
+            data_frame = pd.read_csv(csv_file, index_col='time')
+            print(data_frame)
+
+            print('Writing...')
+
+            start = time.time()
+
+            self._write_client.write("my-bucket", "my-org", record=data_frame,
+                                     data_frame_measurement_name='h2o_feet',
+                                     data_frame_tag_columns=['location'])
+
+            print("Time elapsed: ", (time.time() - start))
+
+        csv_file.close()