According to the survey results , Eight of the top ten most commonly used data tools come from or use Python.Python It is widely used in all fields of data science , Including data analysis , machine learning , Deep learning and data visualization . But you know how to use it Python Do you do data analysis ? What knowledge do you need to learn ? Let's talk about it .
Related to data analysis Python There are many libraries , such as Numpy,pandas,matplotlib,scipy etc. , The operation of data analysis includes data import and export , Data filtering , data description , data processing , statistical analysis , Visualization and so on . Now let's look at how to use it Python Complete data analysis .
Generate data table
There are two common generation methods , The first is to import external data , The second is to write data directly ,Python Supports importing from multiple types of data . At the beginning of use Python You need to import data before you import it pandas library , For convenience , We also import Numpy library . Code is the simplest mode , There are many optional parameter settings , For example, column names , Index column , Data format and so on .
Check data sheet
Python Used in shape Function to view the dimensions of a data table , That is, the number of rows and columns . You can use it info Function to view the overall information of the data table , use dtypes Function to return the data format .Isnull yes Python Functions for testing null values in , You can check the whole data sheet , You can also check the null value of a column separately , The result returned is a logical value , Return with null value True, If not, return False. use unique Function to view unique values , use Values Function to view values in a data table .
Data sheet cleaning
Python The method of dealing with null value is flexible , have access to Dropna Function is used to delete data that contains null values in a data table , It can also be used fillna Function to fill in null values .Python in dtype Is a function to view the data format , Corresponding to it is astype function , Used to change the data format ,Rename Is a function that changes the name of a column ,drop_duplicates Function to remove duplicate values ,replace Function to replace data .
Data preprocessing
Data preprocessing is to sort out the cleaned data for later statistics and analysis , It mainly includes data table merging , sort , Numerical breakdown , Data grouping and marking . stay Python Can be used in merge Function to merge two data tables , The way to merge is inner, In addition, there are left,right and outer mode . use ort_values Function sum sort_index Function to complete the sorting , use where Function to complete data grouping , use split Function implementation .
Data extraction
It mainly uses three functions ：loc,iloc and ix, among loc Function is extracted by label value ,iloc Extract by location ,ix It can be extracted by label and location at the same time . In addition to lifting data by label and location , Data can also be processed according to specific conditions , For example, using loc and isin Use two functions together , Extract the data according to the specified conditions .
Data filtering summary
Python Used in loc Function with filtering conditions to complete the filtering function , coordination sum and
count Functions can also be implemented excel in sumif and countif Functions .Python The main functions used in are groupby and pivot_table.groupby Is a function for categorizing and summarizing , It's easy to use , Just specify the name of the column to group , You can also specify multiple column names at the same time ,groupby
Group by the order in which column names appear .
of course , use Python There are many operations to do data analysis . If you want me to Python Become your career , It's better to choose a major . I suggest you learn from the content , environment , How's the teaching going , Whether to teach face to face , Let's try it on the spot , What suits you is the best .
Technology