Shulou(Shulou.com)06/03 Report--

Indicators and dimensions are the most commonly used terms in data analysis, they are very basic, but very important, often friends do not understand the relationship between them, only master and understand, it is much easier for us to carry out data analysis. Now let's talk about indicators and dimensions.

1. Indicators

Indicator, a unit or method used to measure the degree of development of things, also has a common name on IT, that is, measurement. For example: population, GDP, income, number of users, profit margin, retention rate, coverage and so on. Many companies have their own KPI indicator system, which measures the quality of their business operations through several key indicators.

Indicators need to be summed up, averaged and other summary calculation methods, and need to be summarized in certain prerequisites, such as time, place, scope, that is, we often talk about the statistical caliber and range.

Indicators can be divided into absolute indicators and relative indicators, absolute indicators reflect the size of indicators, such as population, GDP, income, the number of users, while relative indicators are mainly used to reflect the quality of indicators, such as profit margin, retention rate, coverage rate and so on. When we analyze the degree of development of a thing, we can analyze it from the perspectives of quantity and quality, so as to measure the degree of development of things in an all-round way.

As I just said, indicators are used to measure the degree of development of things, which is good or bad, which needs to be compared through different dimensions in order to know whether it is good or bad.

2. Dimension

Dimension: it is a characteristic of a thing or phenomenon, such as gender, region, time, etc. Among them, time is a common and special dimension. through the comparison before and after time, we can know whether the development of things is good or bad. For example, the number of users increased by 10% over the previous month and 20% over the same period last year. This is the comparison of time, also known as vertical comparison.

Another comparison is the horizontal ratio, such as the comparison of population and GDP in different countries, the comparison of income and users in different provinces, and the comparison between different companies and different departments. These are all comparisons between units of the same level, referred to as horizontal ratios.

Dimensions can be divided into qualitative dimensions and quantitative dimensions, that is, according to the data type, the data type is character type (text type) data, that is, qualitative dimension, such as region and gender are both qualitative dimensions. If the data type is numerical data, it is the quantitative dimension, such as income, age, consumption, etc. Generally, we need to do numerical grouping processing for the quantitative dimension, that is, numerical data discretization. The purpose of this is to make the law more obvious, because the finer the grouping is, the less obvious the law is, and finally it becomes the most original pipeline data, then there is no law to follow.

Finally, it is emphasized that only through the two major aspects of the quantity and quality of the development of things, and from the perspective of horizontal ratio and vertical ratio, can we have a comprehensive understanding of the development of things.

To further expand the thinking, I understand it as index separation and dimensional comparison.

In fact, the above ideas can also be referred to in the process of actual product data analysis.

Through the application of a large number of data analysis software tools, it can be found that it mainly includes the following:

Analysis and summary of the overall situation: general situation, changing trend, proportion, etc.

Multi-dimensional analysis: if it is log data, there are already multiple data items, with a data item as the main keyword summary analysis, year-on-year, month-on-month changes, accounting for the total changes. If there is no log data, you need to figure out what is the reason for solving this problem? What data items need to be collected?

Analysis of important scene problems: analyze according to the important issues analyzed and issues of concern to users.

Software and hardware performance management, alarm management, report management, basic parameter configuration and user management, etc.

There are also many commonalities in multi-dimensional analysis, alarm, report, data chart visualization design and presentation, which can be summarized as follows:

1. Is the data presented in a table or a chart? If it is a time range, what is the granularity of time statistics?

two。 What data does the table need to present? The unit of data? How many decimal places do you keep? The method of data calculation? On the basis of sorting?

3. Which chart is used? What is the scope of presentation?

4. Common data item operations: add, delete, modify, query

What are the new required data items? Check repeatability and effectiveness?

Do you need to be reminded to delete? Do you have permission to delete?

What are the modifiable data items? Do you want to verify validity and duplicates after modification? Do you have permission to modify it?

Is the query accurate or fuzzy? Is it a single query or a batch query? Pay attention to the input mode of batch query? What is entered in the query? does it support uppercase and lowercase spaces? Is the query of the data range customized or scoped?

People always think that the technology that has something to do with big data's analysis can only be obtained at a high price. But in fact, big data's analysis of the idea is the most expensive, technology can achieve batch data cleaning, processing, presented faster and more beautiful. But we don't know which data is valid and which data need to be analyzed to get valuable information.

Don't use tactical diligence to disguise strategic laziness

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.