changelog.md

0.1.49

• Fix compatibilities issues with the former Gateway.

0.1.48

• GZIP compression expect a binary input data prior than an utf8 one.
• If Data Warehouse Server can't provide expected reply output format, rename output file to reflect the change.

0.1.47

• Take into account accept parameter to download_edw(), used to select reply output format, accepted format are : application/json, application/parquet, text/csv

0.1.46

• Final compression of CSV file is now optional( compress=True | False )
• Add an extra argument to download_edw() function used to enable transport layer compression ( encoding=gzip|identity )

0.1.45

• Avoid error when killing an Analysis who was already recycled.

0.1.44

• Remove some traces.

0.1.43

• Avoid being connected during JSON -> CSV -> GZ convertions

0.1.42

• fix EDW new hostname

0.1.41

• download_* methods: now check if requested website_name is allowed for current user
• download_edw: dump urllib response if ijson.common.IncompleteJSONError occurs
• download_edw: properly kill job uuid

0.1.40

• Explicit call to action for BlockingIOError in Conn.download_edw
• Solved issue where "p1" column appeared twice in Conn.download_flat_overview_realtime_report

0.1.39

• fix path for SL channel in Conn.download_float_overview_realtime_report
• set correct pandas column type for realtime report and flat overview

0.1.38

• notebook edw_raw_clickview
• typo in filename in Conn.download_edw

0.1.37

• conn.download_flat_realtime_report, incorrectly setting dtype=int64 for floats

0.1.36

• conn.download_realtime_report: fix when date in absent in the response json

0.1.35

• conn.download_datamining: deep copy the object payload to avoid modification on the original object

0.1.34

• New methods for Conn class to retrieve orderpayment and estimatetype

0.1.33

• fix prev version typo

0.1.32

• Support for more channels in flat_download_overview_realtime_report

0.1.31

• new Conn method flat_download_overview_realtime_report
• see notebook

0.1.30

• new Conn method flat_download_realtime_report
• see notebook

0.1.2

• created module for Conn class
• Conn class methods into separate files

0.1.1

• fixed bug in datamining.deduplicate_products where stubnames were not generated correctly

0.1.0

• more unit testing
• cleaned up the Conn class

0.0.99

bug fix, casting payload['view-id'] to str before regex check

0.0.98

Implemented a proper logger for Conn class

0.0.97

pylint review + tests

0.0.96

tests + internal doc/opti

0.0.95

• extra documentation for datamining payload
• new transformation module to work on pandas DataFrame
• new functions in eajupyter
• • button_hide_input_cells
• • remove_class_output_scroll

0.0.94

internal optimizations

0.0.93

fix packaging + update payload

0.0.92

wrapper for datamining payload params

0.0.91

conn.get_website_by_name

0.0.90

filtered out rows without products/marketing touches

0.0.89

deployment workflow mattermost hook secret fix notebook typo

0.0.88

deployment workflow fix

0.0.87

fix strip correctly col_name to push into stubnames array

0.0.86

fix in eaload/generic.py streamline view column name

0.0.85

typo

0.0.84

fix in eaload/generic.py with view-id=0, add viewchannel_X to match other views

0.0.83

fix in eaload.datamining deduplicate_touchpoints and deduplicate_products can be used with either csv or pandas Dataframe

0.0.82

uint dtype to int

0.0.81

fixcgiparam download datamining

0.0.80

fix in eaload.datamining.deduplicate_product_cols_file_2_df

0.0.79

fix edw auth token in download_edw

0.0.78

changes in download_edw do not store the edw access token locally anymore

0.0.77

new download_realtime_report it is a simple API which will evolve in the feature (waiting for backend) notebook

0.0.76

different functions to load csv files into dataframe different functions to create widgets and buttons to interact with jupyter get started notebooks now use internal eulerian name as header (more consistent) pandas column dtype automatically set in eaload functions

0.0.75

fix Conn.download_datamining method day downloaded multiple times various fix

0.0.74

improve jupyter_helper.hide_cells() various fix

0.0.73

fix datamining helper

0.0.72

Conn.download_datamining

• return now an array of files instead of a string
• deletion of params jobrun_id / output_filename
• new param n_slice_days (default=31) split query to datamining by slice new csv_files_2_df method

0.0.71

log http get request in Conn.download_datamining method

0.0.70

Packaging fix

0.0.69

New helper method Jupyter: Add a html button to trigger cell visibility according ot the given tag code

from  eanalytics_api_py import jupyter_helper
jupyter_helper.hide_cells(tag=["code"])

0.0.68

get_view_id_name_map Added default view_id=0

0.0.67

New Conn method get_view_id_name_map

0.0.66

new param status_waiting_seconds for download_edw and download_datamining method

0.0.65

Optimized download datamining with a lot of data datamining_helper

0.0.64

Fix indexing issue in datamining_helper

0.0.63

New module: datamining_helper.py New method: deduplicate_product_cols_file_2_df See here