Data Analysis-Methods & Processes & Tools

• Data analysis refers to the science and art of collecting, processing, and collating data in a targeted manner, and using statistics and mining techniques to analyze and interpret the data.
• From the perspective of the industry, data analysis is a process of collecting, sorting, processing and analyzing data based on a certain industry purpose, and extracting valuable information.

1. Data analysis method

1.1 Comparative analysis method

Comparative analysis is also called comparative analysis. It compares two or more interrelated index data, analyzes its changes, and understands the essential characteristics and development laws of things.
It is mainly reflected in the similarities and differences in the size, level, and speed of the comparison objects.

The form of comparative analysis

According to different types of indicators, the comparison forms include absolute numbers, relative numbers, averages, percentages, and frequencies.

The point to be emphasized here is: when making comparisons, the types of indicators must be unified. The measurement units and calculation methods used must be consistent. Only a unified comparison standard has the meaning of comparison.

Horizontal contrast

• Horizontal comparison is to compare yourself with others, such as industry comparison, product comparison

Vertical contrast

• Vertical comparison is to compare with one of your own indicators, such as: year-on-year, chain-on

1.2 Subdivision analysis method

Group comparison

• Used to continuously subdivide certain data to analyze the data relationship in various subdivision situations and find the root cause that affects the data

Cross-contrast

• Analyze the relationship between two dimensions of an indicator, that is, cross-arrange two related variables and their values ​​in a table at the same time, so that each variable value becomes the intersection of different variables, forming a cross table, so as to analyze the intersection The relationship between the variables in the table.

1.3 A/B test

Develop two solutions for the same goal (for example, two pages), let some users use plan A and the other use plan B, record the user's usage, and see which plan is more in line with the design goal.
Which solution is better for general users, and which design is better? There are many places to test in website design, such as: color series, copy, basic layout, image, title copy, text size, font, etc.

1.4 Funnel analysis method

Conversion funnel analysis is the basic model of business analysis. The most common one is to set the final conversion as the realization of a certain purpose, and the most typical one is to complete the transaction.

2. Data analysis process

2.1 Business perspective

2.3 Engineering perspective

3. Data analysis tools

Application phase	tool
data collection	Crawler: Python; Database: SQL, Hadoop, Hive
data processing	Interface type: Excel, SPSS; Code type: Python, R, SAS; Database: SQL, Hive
Modeling	Interface type: SPSS; Code type: Python, R, SAS
Visualization	Interface type: Excel, PPT, Tabeau; Code type: Python, Echarts, D3.js

3.1 Analysis Tool-Excel

• Knowledge points
  • Basic function
  • Pivot table
  • data visualization
• Mastery
  -Data Processing
  -Data Chart

3.2 Analysis Tool-SQL

• Knowledge points
  • Basic function
  • Pivot table
  • data visualization
• Mastery
  • data processing
  • Data chart

3.3 Tableau & Power BI

• Chart making
• Dashboard construction
• Build a shaped business dashboard

3.4 SPSS

• Knowledge points
  • Descriptive analysis
  • trust level analysis
  • Validity analysis and factor analysis
  • related analysis
  • regression analysis
  • variance analysis
• Mastery
  • Master the SPSS descriptive, statistical analysis, reliability analysis, and consistency analysis of the questionnaire survey
  • Master the relevant analysis methods of SPSS
  • Master ANON hypothesis test analysis

3.5 Python

• Knowledge points
  • scrapy
  • numpy
  • pandas
  • pyEcharts
  • clear
• Mastery
  • Make several complete data projects from data processing to modeling and tuning