Analytics Reports MAIN WIKI

From UG

Revision as of 19:44, 11 October 2011 (edit) Andrei (Talk | contribs) (→Implementation phases) ← Older edit		Revision as of 22:23, 11 October 2011 (edit) (undo) Andrei (Talk | contribs) (→Implementation phases) Newer edit →
Line 91:		Line 91:
	Implementation phases should be the same for every kind of analytics report.		Implementation phases should be the same for every kind of analytics report.
-	=== Phase 1. Prototype of wizard ===	+	=== Phase 1. Prototype of report ===
	* prepare DB,		* prepare DB,
-	* create prototype of report wizard,	+	* create prototype of report,
	* create live "global" table using CT2 data.		* create live "global" table using CT2 data.
-	=== Phase 2. Full wizard feature ===	+	=== Phase 2. Full report ===
-	* create flexible report wizard,	+	* create flexible report,
	* output of all tables and charts.		* output of all tables and charts.

---

Revision as of 22:23, 11 October 2011

Contents 1 Info 2 Core need 3 SA notes (concept vision) 3.1 Main goal 3.2 Data 3.3 Handling 3.4 Output 3.5 Modelling 4 Suggested solution 4.1 Content tasks 4.2 Visualization tasks 5 Implementation phases 5.1 Phase 1. Prototype of report 5.2 Phase 2. Full report

Info

0003003: (Analytics Reports) Create a module to produce this type of reporting

Core need

We want to automate generation of special class of xls reports ("Utilization and Cost Saving Modeling reports"):

• based on CT2 database data
• Complex: mulitab, mix of graphics and tables on one page
• Normally related to utilization rates and costs, times
• Used as a tool for modeling of how to tweak different parameters such as transit times and utilization rates to reduce costs
• Challenge is that every time we need to pull fresh data and apply formulas and charts
• we also want to have “in system” (DR (Dashboard Reports)) version of these reports
• see 2 examples in 0003003
  • examples#1 is well defined in TMS Monthly Stats Report - see Copy of Copy of July Activity Stats Final.xlsx
  • ex#2 - see up+kpi+-+analytics+report+SAMPLE.rar

SA notes (concept vision)

Main goal

Core need:
Automate generation of special class of xls reports ("Utilization and Cost Saving Modeling reports")

That class of reports is designed to give the statistical and operational data to analyze and should help to take the business decision for reduce the cost of freight charges.

Currently there are 2 specific reports that needed to be implemented:

• Elizabeth Arden's TMS Monthly Stats Report. Algorythms of cost savings is based upon reducing the load count to Best possible Load Count.
• Coty report. Algorythm is TBD.

Data

Core need:
Report should be based on CT2 database data

Currently EA monthly statistical report is generated from Descartes TMS in the form of report called "Supplier Detail Previous Month". It contains the summary totals of achieved values (loads, weights, charges etc) for every lane "Supplier-Consignee" per given month. The same lanes with different Carriers are considered as unique.

For both mentioned reports, first task is providing of saving in CT2 database the initial data for reports. So we need to define:

• list/mapping of existing fields that can be used,
• list of new fields that should be added into DB.

Handling

Core need:
Normally related to utilization rates and costs, times
Challenge is that every time we need to pull fresh data and apply formulas and charts

First need get list of indicators for every report. Then:

• develop a procedure of input/export initial data,
• make calculation of needed indicators by given formulas.

Output

Core need:
Complex: mulitab, mix of graphics and tables on one page
We also want to have “in system” (DR (Dashboard Reports)) version of these reports

• System should provide an availability of selecting different output options - for admin and user.
• Every indicator should have its own table or chart (on common or separate page).

Modelling

Core need:
Used as a tool for modelling of how to tweak different parameters such as transit times and utilization rates to reduce costs

This as in http://en.wikipedia.org/wiki/Scientific_modelling.

TBD

Suggested solution

It is suggested to divide all implementation tasks onto two parts: content and visualization.

Content tasks

• Preparing of data:
  • define mapping of existing fields,
  • define list of new filds to add into DB structure.
• Getting of data:
  • develop procedure of input/export data into DB.
• Handling of data:
  • define indicators (given and calculated),
  • specify appropriate calculations.

Visualization tasks

• Configuration of report:
  • admin part - visibility rules, constraints and defaults,
  • set of output columns.
• Output of report:
  • versions of report - standalone/scheduled/dashboard,
  • output format - XLS/HTML,
  • output objects - tables & charts.

Implementation phases

Implementation phases should be the same for every kind of analytics report.

Phase 1. Prototype of report

• prepare DB,
• create prototype of report,
• create live "global" table using CT2 data.

Phase 2. Full report

• create flexible report,
• output of all tables and charts.