当前位置:网站首页>Steps and common methods of data analysis
Steps and common methods of data analysis
2022-07-22 20:50:00 【roo_ one】
Here's the catalog title
The six steps of data analysis
1. Analysis design
The first is to clarify the purpose of data analysis , Only a clear purpose , Data analysis will not deviate from the direction , Otherwise, the data analysis results obtained are not only of no guiding significance , That is, goal guidance
2. data collection
Data collection is based on a defined data analysis framework , The process of collecting relevant data , It provides material and basis for data analysis .
3. Data processing
Data processing refers to processing and sorting the collected data , Form a style suitable for data analysis , Ensure the consistency and effectiveness of data . It's an essential stage before data analysis .
The basic purpose of data processing is to extract from a large number of 、 May be messy 、 Extracting and deriving from incomprehensible data is valuable for solving problems 、 Meaningful data . If the data itself has errors , Then even if the most advanced data analysis methods are adopted , The result is also wrong , It has no reference value , Even misleading decisions .
Data processing mainly includes data cleaning 、 Data transformation 、 Data extraction 、 Data merging 、 Data calculation and other processing methods . General data need to be processed to be used for subsequent data analysis , Even if “ clean ” The original data also needs to be processed before it can be used .
4. Data analysis
Data analysis refers to the use of appropriate analytical methods and tools , Analyze the collected data , Extract valuable information , The process of forming valid conclusions .
At the stage of determining the thinking of data analysis , The data analyst should determine the appropriate data analysis method for the content to be analyzed . At this stage , Can control data , Calmly analyzed and studied .
5. Data presentation
Talk with a chart .
Common data charts include pie charts 、 Bar charts 、 Bar chart 、 Broken line diagram 、 Scatter plot 、 Radar chart, etc , Of course, these charts can be further processed , Make it into the graphics we need , For example, the pyramid 、 Matrix diagram 、 Waterfall Plot 、 Funnel diagram 、 Pareto et al .
6. Report writing
Data analysis report is actually a summary and presentation of the whole data analysis process . Through the report , Analyze the cause of the data 、 The process 、 The results and recommendations are presented in full , For the reference of decision makers . Therefore, the data analysis report is to evaluate the quality of enterprise operation through all-round scientific analysis of data , Science for decision makers 、 Rigorous decision-making basis , In order to reduce the risk of enterprise operation , Improve the core competitiveness of enterprises .
A good analysis report , First of all, there needs to be a good analytical framework , And the level is clear , illustrated , Can let the reader at a glance . The structure is clear 、 Clear priorities can make the reading object correctly understand the content of the report ; illustrated , It can make the data more lively , Improve visual impact , Help readers to be more vivid 、 Look at the problems and conclusions intuitively , And that leads to thinking .
secondly , There needs to be a clear conclusion , An analysis without a definite conclusion is not an analysis , At the same time, it lost the significance of the report , Because at first, the analysis was carried out to find or prove a conclusion , So don't give up the essentials and seek the end .
Third , There must be suggestions or solutions , As a decision maker , It's not just about finding problems , More importantly, suggestions or solutions , So that they can refer to when making decisions . therefore , Data analysts not only need to master data analysis methods , And to know and be familiar with the business , In this way, according to the business problems found , Put forward feasible suggestions or solutions .
Reference resources : The basic steps of data analysis
Common ideas of data analysis
边栏推荐
- Airtest 进行WebUI自动化测试(selenium)
- Charles 抓包原理与实践
- 1045 favorite color stripe (30 points)
- 1049 counting ones (30 points)
- 1034 head of a gang (30 points)
- 面试官:生成订单30分钟未支付,则自动取消,该怎么实现?
- selenium测试框架快速搭建(ui自动化测试)
- 具有非矩形布局1结构的电流镜的梯度灵敏度降低
- 【FPGA】:ip核--DDR3
- Pytoch realizes the fixation of some layers of the network without updating the loopback
猜你喜欢
随机推荐
Common centroid capacitor layout generation considering device matching and parasitic minimization
数据倾斜
1053 path of equal weight (30 points)
初次见面 多多关照
鏈棧實現(C語言)
信号耦合约束下扭曲共质心电容器阵列的布线能力
Configure your own VIM editor God
多线程04--线程的有序性
Matching of MOS transistors with different layout styles
多线程05--Lock
【FPGA】:aurora核的使用
postman接口测试
希尔排序(最小增量排序)
C language (Itoa function)
类的属性新建(初级理解)
多线程03--synchronized和锁升级
具有任意电容比的共质心电容阵列的自动生成
使用多种加权方法在 SAR ADC 中放置二进制加权电容阵列
ThreadLocal遇到线程池出现数据问题和解决方案
Process fork