obsei

---

Pandas DataFrame

---

import pandas as pd
from obsei.source.pandas_source import PandasSource, PandasSourceConfig

# Initialize your Pandas DataFrame from your sources like csv, excel, sql etc
# In following example we are reading csv which have two columns title and text
csv_file = "https://raw.githubusercontent.com/deepset-ai/haystack/master/tutorials/small_generator_dataset.csv"
dataframe = pd.read_csv(csv_file)

# initialize pandas sink config
sink_config = PandasSourceConfig(
   dataframe=dataframe,
   include_columns=["score"],
   text_columns=["name", "degree"],
)

# initialize pandas sink
sink = PandasSource()

Step 2: Configure Analyzer

Note: To run transformers in an offline mode, check transformers offline mode.

Some analyzer support GPU and to utilize pass device parameter. List of possible values of device parameter (default value auto):

1. auto: GPU (cuda:0) will be used if available otherwise CPU will be used
2. cpu: CPU will be used
3. cuda:{id} - GPU will be used with provided CUDA device id

Slack from obsei.sink.slack_sink import SlackSink, SlackSinkConfig # initialize slack sink config sink_config = SlackSinkConfig( # Provide slack bot/app token # For more detail refer https://slack.com/intl/en-de/help/articles/215770388-Create-and-regenerate-API-tokens slack_token="<Slack_app_token>", # To get channel id refer https://stackoverflow.com/questions/40940327/what-is-the-simplest-way-to-find-a-slack-team-id-and-a-channel-id channel_id="C01LRS6CT9Q" ) # initialize slack sink sink = SlackSink()
Pandas DataFrame from pandas import DataFrame from obsei.sink.pandas_sink import PandasSink, PandasSinkConfig # initialize pandas sink config sink_config = PandasSinkConfig( dataframe=DataFrame() ) # initialize pandas sink sink = PandasSink()
2	Observe app reviews from Google play store, PreProcess text via various text cleaning functions, Analyze them by performing text classification, Inform them to Pandas DataFrame and store resultant CSV to Google Drive
	PlayStore Reviews → PreProcessing → Classification → Pandas DataFrame → CSV in Google Drive
3	Observe app reviews from Apple app store, PreProcess text via various text cleaning function, Analyze them by performing text classification, Inform them to Pandas DataFrame and store resultant CSV to Google Drive
	AppStore Reviews → PreProcessing → Classification → Pandas DataFrame → CSV in Google Drive
4	Observe news article from Google news, PreProcess text via various text cleaning function, Analyze them via performing text classification while splitting text in small chunks and later computing final inference using given formula
	Google News → Text Cleaner → Text Splitter → Classification → Inference Aggregator

Tips: Handle large text classification via Obsei

Documentation

For detailed installation instructions, usages and examples, refer to our documentation.

Support and Release Matrix

	Linux	Mac	Windows	Remark
Tests	✅	✅	✅	Low Coverage as difficult to test 3rd party libs
PIP	✅	✅	✅	Fully Supported
Conda				Not Supported

Discussion forum

Discussion about Obsei can be done at community forum

Changelogs

Refer releases for changelogs

Security Issue

For any security issue please contact us via email

Stargazers over time

Maintainers

This project is being maintained by Oraika Technologies. Lalit Pagaria and Girish Patel are maintainers of this project.

License

• Copyright holder: Oraika Technologies
• Overall Apache 2.0 and you can read License file.
• Multiple other secondary permissive or weak copyleft licenses (LGPL, MIT, BSD etc.) for third-party components refer Attribution.
• To make project more commercial friendly, we void third party components which have strong copyleft licenses (GPL, AGPL etc.) into the project.

Attribution

This could not have been possible without these open source softwares.

Contribution

First off, thank you for even considering contributing to this package, every contribution big or small is greatly appreciated. Please refer our Contribution Guideline and Code of Conduct.

Thanks so much to all our contributors