Data collation is often a major pain point for many people in data analysis. However, highly repetitive and frequent reports not only require a lot of time to make, but also often require high timeliness. Therefore, in many cases, even if you know well Analysis can help improve yield, but because analysis results cannot be obtained in time, the use of analysis to help improve production capacity is abandoned. Therefore, if the entire process of data collation and analysis can be integrated into a standard analysis process, the time for making analysis reports can be greatly reduced and the analysis efficiency can be improved.
new version JMP17 Record the analysis process to realize analysis automation
JMP has a very convenient data table organization tool platform. JMP unique JSL The program editing function allows users to arrange standard operating procedures according to their needs, including data collation and subsequent statistical analysis, and completes the production of analysis reports. With the new release, JMP When the user edits and organizes the data table, he can preview the state of the data table after sorting in advance to avoid generating wrong data tables; in addition, JMP improved JSL Log The presentation and structure of the , so that users can edit scripts more quickly and conveniently; or they can use JMP 17 new function workflow builder Complete the formulation of a complete analysis process to achieve the purpose of automation.
From data cleansing to analysis automation, semiconductor CP Data as an example
We utilize semiconductor CP As an illustration, when the wafer is fabricated, the tester will use the The probe card is connected to the test pad on the wafer for electrical testing to confirm the electrical test performance of the wafer. If there is a Die that exceeds the specification, it will be marked, and the subsequent packaging and testing process will not be performed to avoid defective products. .
JMP can quickly analyze the data thrown out by the testing machine to determine which wafers have high defects or test items with poor electrical test performance, helping engineers to make a quick preliminary judgment. JMP supports the import of multiple data formats (Figure 1) and there are many ways to import data.
Figure 1. Data formats supported by JMP for import
When you use open under File to open a text file (.txt, .csv) or an excel format (.csv), you can choose to open it as a preview, and JMP will guide you to the import wizard platform to help users meet their needs To import the required content, letâs take the specification of grabbing CP data as an example. If we want the first column to be the header of the data table, we can check File contains column names on line on the import wizard platform, and in the following methods Fill in the box with 1, which means the data is in the first column.
We can also select the scope of the data. For example, the data specification only includes the third and fourth columns of the original data, then we can mark 3 on the box after Data starts on line, and select lines from start of under subset file, fill the box on the left with 2, which means that the data starts from the third column and continues for two columns, that is, the third and fourth columns containing the original data. (Figure II)
Figure II, use open under import wizard Import data specification content according to requirements
Then we use the same concept to convert CP Import test data, CP The data starts from the twelfth column of the original data, use open turn on CP after the path, in On the import wizard, check File contains column names on line, and fill in 12 in the following box, because all the data below 12 columns need to be selected, just keep the default of the original subset. (Picture 3)
Figure three, use open under import wizard Import as needed CP Test data content
Next we have to organize and clean the data we imported. Using JMP When doing data analysis, we must first organize the data table, so that the format of the data table is that each column contains various information of each sample point, and each row represents a parameter, result or some related information. ( Figure four )
Figure 4, JMP Data table analysis data display format
We can find that the original data does not conform to the JMP format of analysis ( Figure five ) , we hope to organize it into the data type shown in Figure 4. have JMP of table With the help of the function list, such organization becomes very convenient. Open JMP above the data sheet table feature list, found JMP Provides many tools for data table organization, including Summary( summary ) , Subset( extract subgroup ) , Sort( to sort ) , Stack( to stack ) , Split( to split ) , Transpose( Transpose ) , Join( intersection union ) , Update( renew ), concatenate( glue ) etc. Each functional platform includes a preview platform to help users understand what kind of data table will be generated by the current adjustment, avoid making wrong data tables, and save time for purifying data tables.
Figure 5, CP data import JMP initial data sheet for
In this example, we first set the first to fourth Site The data is stacked and retained Parameter data of ( Figure six ) , check stack under the platform Auto Refresh The preview platform can be opened.
Figure 6, Stack Platforms and preview windows
Then the stacked Data Data based on Parameter split and keep Site variables, check the same stack under the platform Auto Refresh Open preview platform ( Figure seven ) .
Figure 7, Split Platforms and preview windows
After completing the collation of the above data sheets, we found that CP The positional parameters of x and Y , which was originally combined into one parameter ( Figure eight ) , we can select this parameter and use Col under the function list Utlities >Text to columns To split the parameter content ( Figure 9 ) .
figure eight, original data sheet X&Y parameter
Figure Nine, use Co>Utlities>Text to columns to split the parameters
At this point, we have completed the sorting and purification of the data.
use JMP Automate your analysis: JSL Action Recorder , JMP17 Workflow Builder
(1) JSL Action Recorder
In order to allow working partners and project collaborators to understand the information brought by the data more quickly, the company may have some fixed-output analysis reports. JMP With many different demand-oriented analysis platforms, the diverse and compatible statistical analysis tools are suitable for users with various statistical analysis needs. use JMP of JSL , you can customize the required platform according to the needs of users, or you can use JSL To edit the overall analysis process, including the aforementioned data collection, collation, cleaning, report production and output, to achieve the goal of reporting automation.
As mentioned earlier, JMP Hope to be closer to the user's usage pattern. Considering that most users have little experience in software editing, JMP continuous improvement JSL mode to help users get started more easily JSL edit. JMP to improve log The structure has changed from the original scattered software language records to the form of modular blocks. JMP call it Action Recorder( Figure ten ) . With this drawer type log structure, users can find what they need more quickly JSL language blocks, accelerated JSL collection, editing and integration.
Figure 9, JMP Modularized form of log , Action recorder
(2) JMP 17 Workflow Builder
Based on this concept, JMP also provides Workflow builder as JMP A powerful assistant for workflow arrangement ( Figure Eleven ) . Compared to Action Recorder , Workflow builder Provides a more visual and physical process arrangement interface , such an interface is convenient for the general public to do more intuitively JMP The process arrangement of various actions, including the increase, decrease and adjustment of process items, combined with the block program structure also facilitates users to understand where problems occur and correct problems.
Figure 10, JMP Workflow Builder
Then demonstrate how to operate the workflow builder . when wanting to construct workflow can be opened when File>New>Workflow , turn on Workflow Builder, Press the record button in the upper left corner ( red circle ) , to start the recording of actions. At this time, a white square will appear in the middle of the red circle when the button is pressed, and the next operation actions will be recorded as work items one by one until the record button is pressed again to return to the red circle pattern. ( Figure 12 ) .
Figure 12, Workflow Builder record key
You can also execute the desired action first, and the action will be recorded in the JMP Log History below On, this record is also recorded synchronously to log window, select JMP Log History In the action you want to integrate, press the up button on the right to integrate the action into Workflow step( Figure 13 ) .
Figure thirteen, Select the action item to move and press the up key to join Workflow step list
For Workflow steps action item, you can choose to move up on the right , Move down or delete button adjustment ( Figure Fourteen ) .
Figure XIV, Select the action item to be moved, and press the move key or delete key to edit Workflow step list
If you want to modify the JSL of the edit action item , or to mark the action, you can open the right step settings , edit and modify the action item ( Figure 15 ) .
Figure fifteen, turn on Step Settings edit modify action item
Press Step next to record The execution button can restore the action item, perform a full action, or move forward or backward step by step. When an action item encounters an error, an error mark will appear on the right side of the action item; if the action is executed smoothly, a tick symbol will appear ( Figure sixteen ) .
Figure sixteen, According to the mark on the right, understand the steps of the problem error
When the work item arrangement is completed, the entire schedule can be saved. Press Workflow Builder under the red triangle create workflow package , you can save the entire process as .jmpflow file and share it directly with other users ( Figure seventeen ) .
Figure seventeen, use Create Workflow Package save schedule
JMP According to the usage habits of users, we will continue to improve and revise to meet the needs of users, and hope that most users can complete the analysis reports they want at their fingertips.
Using Workflow Builder and JMP Table Editing tools, users can make normal work and reports into JMP schedule or JSL , quickly complete the report, save the valuable time of the user, improve work efficiency, apply JMP 17 try out Try it now!
Recommended reading:
