What Is a Book of Original Entry?

First-hand data (original data) refers to the first-hand materials (first-hand data) obtained in all geological surveys. The original data is the basis for comprehensive research and analysis. The original data is the basis for comprehensive research and analysis. For example, the original data in regional geological surveys include: original records and maps of geological observation routes, measured sections, mineral point inspections, geophysical and geochemical exploration work, etc .; various types of rocks, minerals, paleontological specimens, and various sample sampling and identification records; Quantitative measurement (including water sediment measurement), heavy sand analysis, thin sheet analysis and identification results; there are various data, registration cards, registration books, data processing programs and records. The completeness, detail, and reliability of the above data will directly affect the quality of geological maps and the degree of geological and mineral research.

Original data with

First of all, it is necessary to clarify the scope of data collection, such as the geographical scope and time range, and the compilation of professional journals must be limited to the professional scope. The geographical scope is the scope determined according to the administrative area, but attention should be paid to the collection of relevant departments' economic, cultural radiation and radiation data beyond the geographical scope. The time range is the upper and lower limit stipulated in the rules of each book. The scope of professionalism, all majors are within the scope, we should pay attention to breaking the boundaries of the department, do not collect non-professional data outside the majors.

Secondly, we must understand the source of the data, which is extremely extensive. According to the different forms of data, it is generally divided into text data,

Asking questions about the source

Percentiles are important indicators for statistical descriptions. In particular, the median, not only has good stability, but also has a wide range of applications. It can be used for skewed data, data with unknown distribution, and data without a certain value at both ends of the distribution. The methods of calculating the median and percentiles in the frequency table are described in detail in the relevant monographs and textbooks of medical statistics ^[1] , but for the original data, only the calculated median (M) is given. The formula:

M = X _{(n + 1) / 2} , n is an odd number

M = (X _{n / 2} + X _{(n / 2 + 1)} ) / 2, n is an even number (1)

Where n is the total number of observations, ((n + 1) / 2), (n / 2) and (n / 2 + 1) are the ranks of observations in an ordered sequence, and X _{(n +1) / 2} , X _{n / 2} and X _{(n / 2 + 1)} are the observed values at the corresponding rank.

The formula (1) is only applicable to the calculation of the median, and the method of calculating the percentile using the original data has not yet seen relevant reports. This brings a lot of inconvenience to teaching work and practical application. Through research, this paper gives a general formula for calculating percentiles with raw data.

General formula for calculating percentile and median from raw data

When calculating the percentile X _p from the original data, the key is to find the corresponding percentile at the minimum value (the first value in the ordered sequence) and the maximum value (the nth value in the ordered sequence). Between the first and nth digits, and the right side occupies the 1-p part of the entire interval. Therefore, if d represents the rank of the percentile X _p , then

d = 1 + n-1 p

X _p = X ([d]) + (X [d + 1] -X ([d])) (d- [d]) (2)

[D]) in formula (2) represents taking the integer part of d, and X ([d]) and X [d + 1] respectively represent the observed values at [d] and [d + 1]. Applying formula (2) can find the percentile and median of the original data, so formula (2) can replace formula (1). ^[2]