CN112287191A - Model display method and device and electronic equipment - Google Patents

Abstract

The invention discloses a model display method, a model display device and electronic equipment, and belongs to the technical field of data processing. The specific implementation scheme comprises the following steps: receiving model display information input by a user; acquiring target node information in a model visualization file of a target tree model according to the model display information; and performing model display based on the target node information. Therefore, in the process of visually displaying a very complicated forest tree or a relatively large model, the whole model file does not need to be loaded, the model display information input by a user is used, and the corresponding target node information is loaded based on the model visual file, so that the occupation of memory resources is reduced, and the display efficiency is improved.

Description

Model display method and device and electronic equipment

Technical Field

The invention relates to the technical field of data processing, in particular to a model display method and device and electronic equipment.

Background

In the prior art, a tree model is often stored by adopting a JSON file. When the visual display of the tree model is realized, the whole JSON file of the tree model is required to be loaded into the memory at first, and then the JSON file is analyzed, so that the tree model is displayed. In this case, if the tree model is a random forest model or a relatively complex model, and the JSON file itself corresponding to the tree model is relatively large, a relatively large memory is occupied when the corresponding JSON file is loaded and parsed while the digital model is displayed, thereby resulting in low display efficiency.

Disclosure of Invention

The embodiment of the invention aims to provide a model display method, a model display device and electronic equipment, and aims to solve the problem of low display efficiency of the existing tree model display method.

In order to solve the technical problem, the invention is realized as follows:

in a first aspect, an embodiment of the present invention provides a model display method, including:

receiving model display information input by a user;

acquiring target node information in a model visualization file of a target tree model according to the model display information;

and performing model display based on the target node information.

Optionally, the target tree model includes at least one sub-tree model, the model visualization file includes at least one model structure visualization file, and each model structure visualization file corresponds to one sub-tree model in the target tree model;

the obtaining of the target node information in the model visualization file of the target tree model according to the model display information includes:

acquiring target node information in a model structure visualization file of a target sub-tree model according to the model display information;

and the target sub-tree model is a sub-tree model in the target tree model.

Optionally, the number of the target subtree models is multiple; the model display based on the target node information comprises the following steps:

performing paging display on the model according to the target node information in the model structure visualization file of each target sub-tree model;

wherein, a target subtree model is displayed on each page.

Optionally, the model display information includes target tree information, starting node information, and the number of layers to be displayed;

the obtaining of the target node information in the model structure visualization file of the target sub-tree model according to the model display information includes:

determining a model structure visualization file of a target sub-tree model corresponding to the target tree information according to the target tree information;

and selecting the target node information from the model structure visualization file of the target sub-tree model based on the starting node information and the number of layers to be displayed.

Optionally, each node information of the target sub-tree model corresponds to each row of content in the corresponding model structure visualization file;

selecting the target node information from the model structure visualization file of the target sub-tree model based on the starting node information and the number of layers to be displayed, wherein the selecting comprises the following steps:

s1: storing the initial node information into a queue to be found;

s2: reading the content of each row of the model structure visual file of the target sub-tree model, and judging whether the read node information corresponding to the current row is stored in the queue to be found;

s3: when judging that the node information corresponding to the current row is stored in the queue to be found, caching the content of the current row; the content of the current line is the information of the starting node corresponding to the information of the starting node;

s4: determining information for each child node of the starting node based on the content of the current line;

s5: deleting the initial node information from the queue to be found, and storing the information of each child node into the queue to be found;

s6: and repeating the steps from S2 to S5 for each child node until the information of all nodes of the layer number to be displayed is cached.

Optionally, each content in the model structure visualization file corresponds to each node information of the target sub-tree model;

the model display based on the target node information comprises the following steps:

loading the target node information according to rows;

and performing model display based on the loaded node information.

Optionally, the method further includes:

obtaining model information of a trained target model;

and generating a model visualization file of the target model according to the model information.

Optionally, the model information includes at least one of: the number of the sub-tree models, the name of each sub-tree model, the structural information of each sub-tree model and the relationship among the sub-tree models.

Optionally, generating a model visualization file of the target model according to the model information includes:

respectively aiming at each subtree model, the following processes are executed:

traversing the nodes in the sub-tree model by adopting a precedence traversal method;

and storing the structure information of the sub-tree model according to the traversal sequence of the nodes in the sub-tree model to obtain a model structure visualization file of the sub-tree model.

Optionally, after obtaining the target node information in the model visualization file of the target tree model according to the model display information, the method further includes:

and caching the target node information.

Optionally, after receiving the model display information input by the user, the method further includes:

judging whether cached node information comprises target node information matched with the model display information or not according to the model display information;

and when the target node information matched with the model display information is judged to be included, performing model display based on the target node information.

Optionally, the obtaining target node information in a model visualization file of a target tree model according to the model display information includes:

sending a request message to the storage device based on a preset transmission rule; wherein the request message includes the model display information;

and receiving the target node information fed back by the storage device based on a preset transmission rule.

Optionally, the target node information is compressed and transmitted in an array format.

Optionally, the method further includes:

receiving a first input of a user on an end node of a currently displayed subtree model;

responding to the first input, taking the end node as a starting node, and acquiring target node information in a model structure visualization file of the currently displayed subtree model based on the information of the starting node and a preset display layer number;

and updating the currently displayed subtree model based on the target node information.

Optionally, the method further includes:

receiving a second input of the user on the first node of the currently displayed subtree model;

responding to the second input, and judging whether the display layer where the first node is located exceeds a preset display layer;

determining a target node of a decision path of the first node on an Nth display layer under the condition that the display layer where the first node is located exceeds the preset display layer; n is greater than 0 and less than the number of layers of the preset display layer;

and removing part or all of the sibling nodes of the target node and the child nodes of the part or all of the sibling nodes from the currently displayed subtree model.

Optionally, the model structure visualization file is stored in a form of a table or a text.

Optionally, the model visualization file includes a visualization description file, a model structure visualization file, and a tree model visualization cache file;

the visualization description file comprises basic information of the target tree model, the model structure visualization file comprises structure information of the target tree model, and the tree model visualization cache file comprises cache information of the target tree model.

In a second aspect, an embodiment of the present invention provides a model display apparatus, including:

the first receiving module is used for receiving model display information input by a user;

the first acquisition module is used for acquiring target node information in a model visualization file of a target tree model according to the model display information;

and the display module is used for carrying out model display based on the target node information.

Optionally, the target tree model includes at least one sub-tree model, the model visualization file includes at least one model structure visualization file, and each model structure visualization file corresponds to one sub-tree model in the target tree model;

the first obtaining module is specifically configured to:

acquiring target node information in a model structure visualization file of a target sub-tree model according to the model display information;

and the target sub-tree model is a sub-tree model in the target tree model.

Optionally, the number of the target subtree models is multiple; the display module is specifically configured to:

performing paging display on the model according to the target node information in the model structure visualization file of each target sub-tree model;

wherein, a target subtree model is displayed on each page.

Optionally, the model display information includes target tree information, starting node information, and the number of layers to be displayed;

the first obtaining module comprises:

the determining unit is used for determining a model structure visualization file of a target sub-tree model corresponding to the target tree information according to the target tree information;

and the selecting unit is used for selecting the target node information from the model structure visualization file of the target sub-tree model based on the starting node information and the number of layers to be displayed.

Optionally, each node information of the target sub-tree model corresponds to each row of content in the corresponding model structure visualization file;

the selecting unit is specifically configured to perform the following processes:

s1: storing the initial node information into a queue to be found;

s2: reading the content of each row of the model structure visual file of the target sub-tree model, and judging whether the read node information corresponding to the current row is stored in the queue to be found;

s3: when judging that the node information corresponding to the current row is stored in the queue to be found, caching the content of the current row; the content of the current line is the information of the starting node corresponding to the information of the starting node;

s4: determining information for each child node of the starting node based on the content of the current line;

s5: deleting the initial node information from the queue to be found, and storing the information of each child node into the queue to be found;

s6: and repeating the steps from S2 to S5 for each child node until the information of all nodes of the layer number to be displayed is cached.

Optionally, each content in the model structure visualization file corresponds to each node information of the target sub-tree model;

the display module comprises:

the loading unit is used for loading the target node information according to rows;

and the display unit is used for carrying out model display based on the loaded node information.

Optionally, the apparatus further comprises:

the second acquisition module is used for acquiring model information of the trained target model;

and the generating module is used for generating a model visualization file of the target model according to the model information.

Optionally, the model information includes at least one of: the number of the sub-tree models, the name of each sub-tree model, the structural information of each sub-tree model and the relationship among the sub-tree models.

Optionally, the apparatus further comprises:

an execution module, configured to execute the following processes for each sub-tree model respectively:

traversing the nodes in the sub-tree model by adopting a precedence traversal method;

and storing the structure information of the sub-tree model according to the traversal sequence of the nodes in the sub-tree model to obtain a model structure visualization file of the sub-tree model.

Optionally, the apparatus further comprises:

and the cache module is used for caching the target node information.

Optionally, the apparatus further comprises:

the first judgment module is used for judging whether the cached node information comprises target node information matched with the model display information or not according to the model display information;

the display module is further configured to: and when the target node information matched with the model display information is judged to be included, performing model display based on the target node information.

Optionally, the first obtaining module includes:

a sending unit, configured to send a request message to a storage device based on a preset transmission rule; wherein the request message includes the model display information;

and the receiving unit is used for receiving the target node information fed back by the storage device based on a preset transmission rule.

Optionally, the target node information is compressed and transmitted in an array format.

Optionally, the apparatus further comprises:

the second receiving module is used for receiving a first input of a user on an end node of the currently displayed subtree model;

a third obtaining module, configured to, in response to the first input, take the end node as a start node, and obtain, based on information of the start node and a preset number of display layers, target node information in a model structure visualization file of the currently displayed subtree model;

and the updating module is used for updating the currently displayed sub-tree model based on the target node information.

Optionally, the apparatus further comprises:

the third receiving module is used for receiving a second input of the user on the first node of the currently displayed subtree model;

the second judgment module is used for responding to the second input and judging whether the display layer where the first node is located exceeds a preset display layer or not;

the determining module is used for determining a target node of a decision path of the first node on an Nth display layer under the condition that the display layer where the first node is located exceeds the preset display layer; n is greater than 0 and less than the number of layers of the preset display layer;

and the removing module is used for removing part or all of the brother nodes of the target node and the child nodes of the part or all of the brother nodes from the currently displayed subtree model.

Optionally, the model visualization file is stored in a form of a table or text.

Optionally, the model visualization file includes a visualization description file, a model structure visualization file, and a tree model visualization cache file;

the visualization description file comprises basic information of the target tree model, the model structure visualization file comprises structure information of the target tree model, and the tree model visualization cache file comprises cache information of the target tree model.

In a third aspect, an embodiment of the present invention provides an electronic device, which includes a processor, a memory, and a program or an instruction stored in the memory and executable on the processor, where the program or the instruction, when executed by the processor, implements the steps of the model displaying method described above.

In a fourth aspect, an embodiment of the present invention provides a readable storage medium, on which a program or instructions are stored, and when the program or instructions are executed by a processor, the program or instructions implement the steps of the model displaying method described above.

In the embodiment of the present invention, after receiving the model display information input by the user, the target node information in the model visualization file of the target tree model may be obtained according to the model display information, and the model display may be performed based on the target node information. Therefore, in the process of visually displaying a very complicated forest tree or a relatively large model, the whole model file does not need to be loaded, the model display information input by a user is used, and the corresponding target node information is loaded based on the model visual file, so that the occupation of memory resources is reduced, and the display efficiency is improved.

Drawings

FIG. 1 is a flow chart of a model display method according to an embodiment of the present invention;

FIG. 2 is a diagram illustrating a tree model to be stored in accordance with an embodiment of the present invention;

FIG. 3 is a schematic diagram of a numerical model in an embodiment of the present invention;

FIG. 4 is a flowchart of obtaining target node information in an embodiment of the present invention;

FIG. 5 is a diagram illustrating a request message in accordance with an embodiment of the present invention;

FIG. 6 is a diagram illustrating a feedback message according to an embodiment of the present invention;

FIG. 7 is a diagram of a decision tree model in an embodiment of the present invention;

FIG. 8 is a schematic diagram of a decision tree model shown in an embodiment of the present invention;

fig. 9 is a schematic structural diagram of a model display apparatus according to an embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

The terms first, second and the like in the description and in the claims of the present invention are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that embodiments of the invention may be practiced other than those illustrated or described herein, and that the objects identified as "first," "second," etc. are generally a class of objects and do not limit the number of objects, e.g., a first object may be one or more. In addition, "and/or" in the specification and claims means at least one of connected objects, a character "/" generally means that a preceding and succeeding related objects are in an "or" relationship.

In order to facilitate understanding of the embodiments of the present invention, the following is first explained.

Currently, the types of models supported by the data analysis system include, but are not limited to, neural network models, forest models, trees (decision tree models), machine learning models, deep learning models, feature selection models, regression models, and the like.

The scheme of the invention is mainly used for realizing the visual display of the tree model. The tree model in the present invention includes, but is not limited to, at least one of: tree models such as forest model, tree model, and the like.

The model display method provided by the embodiment of the invention is described in detail by specific embodiments and application scenarios thereof with reference to the accompanying drawings.

Referring to fig. 1, fig. 1 is a flowchart of a model displaying method according to an embodiment of the present invention, as shown in fig. 1, the method includes the following steps:

step 101: model display information input by a user is received.

Optionally, the model display information may include target tree information, start node information, number of layers to be displayed, and the like. The target tree information may be a number, name or other identification of the target tree, etc. The start node information may be a number, name or other identification of the start node, etc. The number of layers to be displayed may be set by a user or a default number of display layers of the system, such as 5 layers, 6 layers, and the like.

Step 102: and acquiring target node information in a model visualization file of the target tree model according to the model display information.

Step 103: and performing model display based on the target node information.

According to the model display method provided by the embodiment of the invention, after the model display information input by the user is received, the target node information in the model visualization file of the target tree model can be obtained according to the model display information, and the model display is carried out based on the target node information. Therefore, in the process of visually displaying a very complicated forest tree or a relatively large model, the whole model file does not need to be loaded, the model display information input by a user is used, and the corresponding target node information is loaded based on the model visual file, so that the occupation of memory resources is reduced, and the display efficiency is improved.

In the embodiment of the invention, the model visualization file of the tree model is used for storing the meta information of the tree model. The meta information includes basic information of the tree model, structural information of the tree model, index information of the tree model, cache information of the tree model, and the like.

Optionally, the model visualization file may include: visual description files, model structure visual files, tree model visual cache files and the like. The visual description file comprises basic information of the tree model, the model structure visual file comprises structure information of the tree model, and the tree model visual cache file comprises cache information of the tree model.

For example, if a trained tree model is a forest model that includes 100 subtrees, it may be composed of the following files in table 1:

TABLE 1

Numbering	Name (R)	Description of the invention
			1	explanation.meta	Visual description file
2	<treeId>.expl	Model structure visualization file
			3	<treeId>.<nodeId>.cache	Tree model visual cache file
……	……	……

In the embodiment of the present invention, the visualization description file includes basic information of the tree model, where the basic information includes: the name of the tree model, the type of the tree model, and information of all sub-tree models contained in the tree model, etc. Preferably, the visual description file may be in JSON format, as shown below:

optionally, the model structure visualization file includes structure information of the tree model, that is, decision path information of the tree model. One model structure visualization file corresponds to one tree model or sub-tree model. For example, if the stored tree model is a forest model, the number of model structure visualization files included in the model visualization file of the forest model is the same as the number of subtrees included in the forest model, that is, one model structure visualization file stores structure information of one subtree model.

Optionally, the tree model visualization cache file includes cache information of the tree model. Initially, the tree model visual cache file is an empty file, in the subsequent display implementation process, relevant information can be retrieved and displayed based on model display requirements set by a user, and the retrieved relevant information can be stored in the tree model visual cache file.

In an embodiment of the present invention, the target tree model in step 102 may include at least one sub-tree model, where the model visualization file of the target tree model includes at least one model structure visualization file, and each model structure visualization file corresponds to one sub-tree model in the target tree model. Correspondingly, the process of acquiring the target node information in step 102 may include: acquiring target node information in a model structure visualization file of a target sub-tree model according to the model display information; and the target sub-tree model is a sub-tree model in the target tree model.

Further, if the number of the target sub-tree models is multiple, the process of performing model display based on the target node information in step 103 may include: performing paging display on the model according to target node information in the model structure visualization file of each target sub-tree model; wherein, a target subtree model is displayed on each page. Therefore, the display effect can be improved by displaying the sub-tree model in a paging mode, and the problems that one-time display cannot be clearly achieved in one page due to the fact that the structure of the tree model is complex, and performance is poor due to too much rendered data such as background processing, network transmission and front-end rendering are solved.

Optionally, in this embodiment of the present invention, each line of content in the model structure visualization file of the target sub-tree model may correspond to each node information of the target sub-tree model, and correspondingly, the process of performing model display based on the target node information in step 103 may include: and loading the target node information according to rows, and performing model display based on the loaded node information. Therefore, when the subtree model is visually displayed, virtual loading can be performed, node information is read in lines and loaded into a memory based on requirements, and a model visual file of the whole model does not need to be loaded, so that the display effect is improved, for example, an 80M model visual file (the size of the model visual file) is loaded in a line and a line according to requirements, and the memory only needs to occupy several megabytes.

In the embodiment of the invention, a complex tree model can be stored based on the idea of hierarchical paging display. Paging refers to displaying a separate tree model per page, for example, for a forest model, paging is performed based on a subtree model in the forest model. The layering refers to layering based on the layer to which each node in the tree model belongs. Under the guidance of the idea of hierarchical paging display, the specific step of storing a complex tree model is as follows: each sub-tree model corresponds to a model structure visualization file, and each node in the sub-tree model corresponds to a line of content in the model structure visualization file. The content of a line in the model structure visualization file is node information of a corresponding node, and the node information includes but is not limited to: a node number (ID) of the current node, a node name (name), child node information of the current node, a policy of the node (e.g., a decision rule or a decision condition from a parent node to a child node), etc.

Optionally, the model display method in this embodiment may further include: obtaining model information of a trained target model; and generating a model visualization file of the target model according to the model information. Wherein the model information includes at least one of: the number of the sub-tree models, the name of each sub-tree model, the structural information of each sub-tree model and the relationship among the sub-tree models.

Further, the model visualization file includes a visualization description file, generating the model visualization file of the target model includes generating the visualization description file, and the process of generating the visualization description file may include: and generating a visual description file according to the number of the sub-tree models in the model information, the name of each sub-tree model, the type of each sub-tree model and the like.

Further, for each sub-tree model, the process of generating the model structure visualization file may include: traversing the nodes in the subtree model by adopting a precedence traversal method; and storing the structure information of the sub-tree model according to the traversal sequence of the nodes in the sub-tree model to obtain a model structure visualization file of the sub-tree model.

That is, the implementation process of storing the tree model as the model visualization file may be: firstly, traversing the tree model to obtain model information of the tree model; the model information at least includes the number of sub-tree models, the structural information of each sub-tree model (such as the node information of all nodes in the sub-tree model), the relationship between the sub-tree models, and the like; then, based on the principle of hierarchical paging and the obtained model information, a model structure visualization file of each sub-tree model is respectively generated, that is, based on the structure of the sub-tree model, the node information of each node in the sub-tree model is stored in the corresponding row of the model structure visualization file.

In one embodiment, the model structure visualization file of each sub-tree model is generated by a precedence traversal method, that is, according to the top-down order, the first right, the second left principle, and the like of the sub-tree model structure.

In another embodiment, the model structure visualization file of each sub-tree model is generated by adopting a precedence traversal method, namely according to the top-down sequence of the sub-tree model structure and the principle of first left, second right.

The following describes a specific process of storing by using a pre-sequence traversal method in the embodiment of the present invention with reference to fig. 2. As shown in fig. 2, the tree model to be stored has the following specific procedure:

s1: the node 1 is pushed;

s2: the node 1 is popped (the top element is popped), the top element 1 is output, and the left and right child nodes of the node 1 are pushed; where the right child node 4 is pushed first and then the left child node 2 is pushed. At this time, the access to the left child node is in the order of the right child node, and then the first access to the stack is performed.

S3: the node 2 is popped (the top element is popped), the top element 2 is output, and the left and right child nodes of the node 2 are pushed; similarly, the right child node 5 is pushed first, and the left child node 3 is pushed later.

S4: and (4) popping the node 3 (popping the top element), outputting the top element 3, wherein the node 3 is a leaf node, and the leaf node is a node without a child node and is not popped.

S5: the node 5 is popped (the top element is popped), and the top element 5 is output.

S6: and the node 4 is popped (the last stack top element is popped), the stack is empty at this time, and the traversal is finished.

Note that fig. 2 corresponds to traversal results of 1,2,3,5, and 4. In the traversal process, the output stack can be understood as being saved to the model visualization file. And then, based on the traversal result, storing the information of the corresponding tree model to obtain a model structure visualization file.

Optionally, the model structure visualization file in this embodiment may be in any of the following formats: tables, text, etc., i.e., stored in the form of tables, text, etc. Further, the embodiment of the invention can support the user to define the format of the tree structure visual file, and before the model visual file is generated, the user can set the format of the generated model structure visual file.

For example, by traversing the tree model shown in FIG. 3, the final generated model structure visualization file may be as shown in Table 2 below.

TABLE 2

id	name	leftChildId	rightChildId	LeftEdge	rightEdge
						Node numbering	Node name	Left subtree numbering	Right subtree numbering	Left information	Information on the right
0	x3	1	2	<＝0.292	>0.292
						1	1.436
2	X0	3	4	<＝0.181	>0.181
						3	X3	5	6	<＝0.521	>0.521
5	-0.909
						6	-0.878
4	X3	7	8	<＝0.604	>0.604
						7	-0.909
8	-0.912

If the user definition is stored in a text form, one text corresponds to one tree, one line in the text corresponds to one node of the tree, and different information of the same node is divided by preset characters, for example, different information of the same node is divided. By traversing the tree model shown in FIG. 3, the final generated model visualization file may be as follows:

0&x3&1&2&<＝0.292&>0.292

1&1.436&&&&

2&x0&3&4&<＝0.181&>0.181

3&x3&5&6&<＝0.521&>0.521

5&-0.909&&&&

6&-0.878&&&&

4&x3&7&8&<＝0.604&>0.604

7&-0.909&&&&

8&-0.912&&&&

further, in the process of storing the model visualization file, the model visualization file may be compressed, for example, in a zip or rar form.

Furthermore, in order to facilitate subsequent retrieval and search, an index can be established in the process of storing the model visualization file. And the established index may include an index identification and an access address of the file corresponding to the index identification. Specifically, the index may be established based on the identification ID, name, etc. of the tree model. For example, a random forest model Tree with an ID of T includes three subtrees a, b and c, the IDs of the three subtrees are Ta, Tb and Tc, and two-level indexes can be established based on the ID of the random forest model and the IDs of the subtrees.

Optionally, in this embodiment of the present invention, the model display information may include target tree information, start node information, and a number of layers to be displayed. Correspondingly, the process of obtaining the target node information in the model structure visualization file of the target sub-tree model according to the model display information may include: firstly, according to the target tree information, determining a model structure visualization file of a target sub-tree model corresponding to the target tree information; and then, based on the initial node information and the number of layers to be displayed, selecting the target node information from the model structure visualization file of the target sub-tree model.

Further, since the model visualization file is stored in a pre-ordered manner, and the child nodes appear behind the parent node, in this embodiment, when the target node is retrieved, the parent node to be retrieved may be first placed in the queue to be retrieved, if the parent node has left and right child nodes, the left and right child nodes are stored in the queue to be retrieved, the parent node is dequeued, and then the retrieval is continued based on the queue to be retrieved until the nodes with the required number of layers to be displayed are all found, and then the retrieval result may be returned. Therefore, the above process of selecting the target node information from the model structure visualization file of the target sub-tree model based on the starting node information and the number of layers to be displayed may include:

s1: storing the initial node information into a queue to be found;

s2: reading the content of each row of the model structure visual file of the target sub-tree model, and judging whether the read node information corresponding to the current row is stored in the queue to be found;

s3: when judging that the node information corresponding to the current row is stored in the queue to be found, caching the content of the current row; the content of the current line is the information of the starting node corresponding to the information of the starting node;

s4: determining information for each child node of the starting node based on the content of the current line;

s5: deleting the initial node information from the queue to be found, and storing the information of each child node into the queue to be found;

s6: and repeating the steps from S2 to S5 for each child node until the information of all nodes of the layer number to be displayed is cached.

Optionally, in the embodiment of the present invention, after the target node information is obtained, the target node information may be cached, so that when performing model display again, the cached result is directly applied without performing retrieval again, thereby saving retrieval time, and thus implementing quick display and improving the display process.

Further, after receiving model display information input by a user, judging whether cached node information includes target node information matched with the model display information according to the model display information; and when the target node information matched with the model display information is judged to be included, performing model display based on the target node information. And if the cached node information does not comprise the target node information matched with the model display information, acquiring the target node information in the model visualization file of the target tree model based on the model display information so as to display the model.

In a specific embodiment, the target tree information may include a tree model index and a subtree model index.

For example, taking the received model exhibition information including the target tree number, the starting node number, and the number of layers to be exhibited (such as 5 layers, 6 layers, etc.) as an example, as shown in fig. 4, the process of obtaining the required target node information may include:

step 401: and obtaining model display information such as a target tree number treeid, a starting node number nodeid and a level of a layer to be displayed according to the input of a user.

Step 402: and judging whether target node information corresponding to the treeid, nodeid and level exists in the tree model visual cache file.

Step 403: and if so, acquiring corresponding target node information.

Step 404: and if not, acquiring a model structure visualization file corresponding to the treeid according to the treeid.

Step 405: a given nodeid is stored in a queue to be found (NQ).

Step 406: reading the content of each line of the model structure visual file, and judging whether the node number corresponding to the read current line is stored in the NQ.

Step 407: if not, the next row is read.

Step 408: if the node information is stored in the NQ, the node information of the current line is cached in the list.

Step 409: and judging whether the NQ is empty or not, or judging whether all nodes of the layer to be displayed are found.

Step 410: and if so, feeding back all node information of the cache.

Step 411: if not, if the current node to be found has left and right child nodes, the left and right child nodes are stored in the NQ, and the current node to be found is moved out of the NQ.

And then, repeatedly executing the steps 406 to 411, namely, continuously reading the information of the model structure visualization file, and judging whether the node number corresponding to the read current row is in the NQ queue or not until the information of all nodes of the layer to be displayed is obtained.

In the embodiment of the present invention, when obtaining the target node information, a request message may be sent to the storage device based on a preset transmission rule, where the request message includes the model display information, and then the target node information fed back by the storage device based on the preset transmission rule is received.

Alternatively, the preset transmission rule may be a transmission protocol supporting file compression, such as an HTTP protocol supporting zip file compression.

Specifically, before communication, both transmitting parties may stipulate transmission rules (transmission protocols) that both transmitting parties must comply with in data transmission, such as the agreed parameters to be transmitted and the format of each parameter, before transmission, so as to perform transmission based on the agreed transmission rules. In order to be able to transmit normally, the transmitted data needs to satisfy the transmission rule. Therefore, as the transmission protocol already appoints the transmission parameters, the target node information can be compressed and transmitted in an array format, namely, the value of each field is actually transmitted, but the meaning (name) of each field is not transmitted, so that the purposes of compressing the transmission content, improving the transmission speed and improving the performance are achieved to the maximum extent.

For example, as shown in fig. 5 and fig. 6, fig. 5 is a schematic diagram of a request message including model presentation information, and fig. 6 is target node information fed back by the request message shown in fig. 5, where rows [0] [0] represents a node id, rows [0] [1] represents a node name, rows [0] [2] represents a parent node number, and rows [0] [3] represents side information.

Optionally, the model display method in the present application may further include:

receiving a first input of a user on an end node of a currently displayed subtree model;

responding to the first input, taking the end node as a starting node, and obtaining target node information in a model structure visualization file of the currently displayed subtree model based on information of the starting node and a preset display layer number (such as 5 layers, 6 layers and the like);

and updating the currently displayed subtree model based on the target node information.

It should be pointed out, when the depth of the tree model is relatively deep, the page is not placed after one-time display, and too much rendered data can cause poor performances such as background processing, network transmission and front-end rendering, therefore, the application supports the hierarchical local expansion of the tree model, a user can set the number of nodes (default to 5 layers) of one-time display, when the sub-tree model of the nodes which are not expanded is desired to be further checked, 5 layers can be loaded again, the hierarchical display is realized, the problem that too much rendered data can cause poor performances such as background processing, network transmission and front-end rendering is solved, and the whole tree model is displayed once.

In addition, the terminal node of the currently displayed sub-tree model is used as the starting node, and corresponding node information can be further loaded to update the sub-tree model on the basis of the currently displayed sub-tree model, so that the page rendering performance is improved, the phenomena of frame loss and blocking when the updating data volume is large are avoided, and the user use process is smoother.

In one embodiment, when a user clicks an end leaf node of a sub-tree model displayed on a current page, a number (nodeid) of the current end leaf node may be obtained, and information of a child node within a layer 5 of the current end leaf node is obtained from a storage device through an HTTP request; after the information is obtained, based on the number of the current terminal leaf node, the node information of the sub-tree model displayed on the current page is updated by searching the minimum change unit, so that the sub-tree model is updated by the difference, and the re-rendering is performed. Specifically, after response data is taken through HTTP, the node information of the subtree model displayed on the current page can be updated based on the concept of React Virtual DOM (namely, in React, each DOM object has a corresponding Virtual DOM object and a React Diff algorithm concept), the rendering performance of the page is improved by searching for the minimum change unit, the phenomenon of frame loss and pause when the updating data volume is large is avoided, and the using process of a user is smoother.

Optionally, in order to reduce the memory usage and further manage the front-end memory, and improve the user experience, the model display method in the application may further include:

receiving a second input of the user on the first node of the currently displayed subtree model;

responding to the second input, and judging whether the display layer where the first node is located exceeds a preset display layer; the preset display layer can be preset based on actual requirements, such as preset to 15 layers;

determining a target node of a decision path of the first node on an Nth display layer under the condition that the display layer where the first node is located exceeds the preset display layer; n is greater than 0 and less than the number of layers of the preset display layer; such as N equal to 5;

and removing part or all of the sibling nodes of the target node and the child nodes of the part or all of the sibling nodes from the currently displayed subtree model.

For example, for a currently expanded subtree model, a display layer where a currently clicked node is located may be determined through recursive computation, and if the display layer where the clicked node is located exceeds a preset display layer, such as 15 layers, a target node where a decision path of the clicked node is located in an nth (for example, 5) display layer may be determined, and from the currently expanded subtree model, part or all of sibling nodes of the target node and child nodes of the part or all of sibling nodes are removed, that is, the nth display layer is taken as a reference line, and part or all of decision path branches except the target node are deleted, so as to reduce the number of display nodes of the subtree model, thereby reducing a memory occupation situation, further managing a front-end memory, and improving user experience.

In addition, the method and the device can mount required displayed basic information for each single node, and can display details of the decision rule of the node. For example, when a user clicks on a node, the decision path from the initial root node to the node may be presented in a presentation interface.

In order to facilitate the user to know the model structure quickly and accurately, as shown in fig. 3, the visualization interface when the model is displayed may have the following features:

1) forest display form: adopting a paging mode, wherein each page corresponds to a sub-tree model to integrally form a forest model;

2) using a box to represent a node of a subtree model;

3) placing the name of the node inside a box representing the node;

4) marking information on the edge by using edge information (decision strategy) when connecting from a father node to a child node;

5) the left and right child nodes are distinguished in different ways, such as different box colors, different box fill contents, and the like.

The following describes an embodiment of the present invention by taking a power grid system in a specific application scenario as an example.

Specifically, the model display method is applied to the power grid system, and the problem in power grid system simulation load flow calculation can be effectively solved through efficient visual model display. For example, in the power flow calculation of the power grid simulation, when the power flow calculation result of the existing input data does not meet the target requirement, the user may first set the current input data and the power flow calculation target through the user interface, wherein the input data includes initial calculation data for power flow calculation and an initial calculation result corresponding to the initial calculation data. Wherein the calculation target is used for screening the calculation model used for executing the load flow calculation. Specifically, the power flow calculation of the power grid system includes various calculation targets, for example, the execution target of the power flow calculation includes at least one of the following: whether the power flow is converged, power flow voltage adjustment, power flow section power adjustment, power flow direct current line power adjustment and the like. Different computational objectives may be computed using different types of artificial intelligence models. For example, whether the power flow converges or not is determined by calculating through a tree model, and power adjustment of the power flow section is calculated through a CNN model. And secondly, selecting an artificial intelligence model for power grid load flow calculation according to a calculation target input by a user, and processing input data by using the selected artificial intelligence model to obtain a processing result. And then, inputting the processing result into a load flow calculation program to carry out load flow calculation to obtain a new load flow calculation result, judging whether the new calculation result reaches a target value, if so, finishing, and otherwise, continuously processing the input data by using an artificial intelligent model until the calculation result of the load flow calculation reaches the target value.

The application of the visualization model in the power grid load flow calculation is described below by taking the reason of load flow non-convergence in the load flow calculation as an example.

Assuming that the calculation target set by the user is the cause of the non-convergence of the power flow calculation, the power grid power flow calculation system may invoke a decision tree model as described in fig. 7 below for analysis based on the calculation target. Specifically, before calling the decision tree model to perform the power flow analysis, the system needs to construct and train a model for predicting the cause of non-convergence of power flow calculation, obtain the following decision tree model after the model training is completed, and then store the decision tree model into the system based on the storage process of the tree model in the application; after the system determines to use the decision tree model based on a calculation target set by a user, the system analyzes input data by using the decision tree model to determine the reason of the non-convergence of the power flow caused by using the input data.

In order to enable a user to know the reason of the non-convergence of the power flow, the system can display the decision tree model in a user interface, so that the user can clearly know the decision process of the reason of the non-convergence of the power flow, and the user can accurately correct the input data based on the decision process, thereby solving the problem of the non-convergence of the power flow in the power grid simulation system. Furthermore, in the power grid load flow calculation, a plurality of factors are caused to influence load flow convergence, and the analysis process of load flow non-convergence is complex, so that a decision tree model for analyzing the reasons of load flow non-convergence is relatively complex, and in order to efficiently display in effective resources, a user can quickly and accurately know the decision process of the reasons of load flow non-convergence and know the reasons of load flow non-convergence, before the decision tree model is displayed, the user can set the number of layers to be displayed, the displayed initial node and the like through a user interface to display configuration information, so that the system obtains the visual information of the model to be displayed and renders the visual information based on the display configuration information, and displays the specified number of decision layers of the decision tree model in the user interface. For example, a user may set a displayed start node as "whether to calculate a data error" through a user interface, the number of layers to be displayed is 3, and a decision tree model displayed in the user interface is shown in fig. 8.

Referring to fig. 9, fig. 9 is a schematic structural diagram of a model display apparatus according to an embodiment of the present invention, the apparatus is applied to an electronic device, and as shown in fig. 9, the model display apparatus 90 includes:

a first receiving module 91, configured to receive model display information input by a user;

a first obtaining module 92, configured to obtain, according to the model display information, target node information in a model visualization file of a target tree model;

and the display module 93 is configured to perform model display based on the target node information.

Optionally, the target tree model includes at least one sub-tree model, the model visualization file includes at least one model structure visualization file, and each model structure visualization file corresponds to one sub-tree model in the target tree model;

the first obtaining module 92 is specifically configured to:

acquiring target node information in a model structure visualization file of a target sub-tree model according to the model display information;

and the target sub-tree model is a sub-tree model in the target tree model.

Optionally, the number of the target subtree models is multiple; the display module 93 is specifically configured to:

performing paging display on the model according to the target node information in the model structure visualization file of each target sub-tree model;

wherein, a target subtree model is displayed on each page.

Optionally, the model display information includes target tree information, starting node information, and the number of layers to be displayed;

the first obtaining module 92 includes:

the determining unit is used for determining a model structure visualization file of a target sub-tree model corresponding to the target tree information according to the target tree information;

and the selecting unit is used for selecting the target node information from the model structure visualization file of the target sub-tree model based on the starting node information and the number of layers to be displayed.

Optionally, each node information of the target sub-tree model corresponds to each row of content in the corresponding model structure visualization file;

the selecting unit is specifically configured to perform the following processes:

s1: storing the initial node information into a queue to be found;

s2: reading the content of each row of the model structure visual file of the target sub-tree model, and judging whether the read node information corresponding to the current row is stored in the queue to be found;

s3: when judging that the node information corresponding to the current row is stored in the queue to be found, caching the content of the current row; the content of the current line is the information of the starting node corresponding to the information of the starting node;

s4: determining information for each child node of the starting node based on the content of the current line;

s5: deleting the initial node information from the queue to be found, and storing the information of each child node into the queue to be found;

s6: and repeating the steps from S2 to S5 for each child node until the information of all nodes of the layer number to be displayed is cached.

Optionally, each content in the model structure visualization file corresponds to each node information of the target sub-tree model;

the display module 93 includes:

the loading unit is used for loading the target node information according to rows;

and the display unit is used for carrying out model display based on the loaded node information.

Optionally, the apparatus further comprises:

the second acquisition module is used for acquiring model information of the trained target model;

and the generating module is used for generating a model visualization file of the target model according to the model information.

Optionally, the model information includes at least one of: the number of the sub-tree models, the name of each sub-tree model, the structural information of each sub-tree model and the relationship among the sub-tree models.

Optionally, the apparatus further comprises:

an execution module, configured to execute the following processes for each sub-tree model respectively:

traversing the nodes in the sub-tree model by adopting a precedence traversal method;

and storing the structure information of the sub-tree model according to the traversal sequence of the nodes in the sub-tree model to obtain a model structure visualization file of the sub-tree model.

Optionally, the apparatus further comprises:

and the cache module is used for caching the target node information.

Optionally, the apparatus further comprises:

the first judgment module is used for judging whether the cached node information comprises target node information matched with the model display information or not according to the model display information;

the display module 93 is further configured to: and when the target node information matched with the model display information is judged to be included, performing model display based on the target node information.

Optionally, the first obtaining module 92 includes:

a sending unit, configured to send a request message to a storage device based on a preset transmission rule; wherein the request message includes the model display information;

and the receiving unit is used for receiving the target node information fed back by the storage device based on a preset transmission rule.

Optionally, the target node information is compressed and transmitted in an array format.

Optionally, the apparatus further comprises:

the second receiving module is used for receiving a first input of a user on an end node of the currently displayed subtree model;

a third obtaining module, configured to, in response to the first input, take the end node as a start node, and obtain, based on information of the start node and a preset number of display layers, target node information in a model structure visualization file of the currently displayed subtree model;

and the updating module is used for updating the currently displayed sub-tree model based on the target node information.

Optionally, the apparatus further comprises:

the third receiving module is used for receiving a second input of the user on the first node of the currently displayed subtree model;

the second judgment module is used for responding to the second input and judging whether the display layer where the first node is located exceeds a preset display layer or not;

the determining module is used for determining a target node of a decision path of the first node on an Nth display layer under the condition that the display layer where the first node is located exceeds the preset display layer; n is greater than 0 and less than the number of layers of the preset display layer;

and the removing module is used for removing part or all of the brother nodes of the target node and the child nodes of the part or all of the brother nodes from the currently displayed subtree model.

Optionally, the model visualization file is stored in a form of a table or text.

Optionally, the model visualization file includes a visualization description file, a model structure visualization file, and a tree model visualization cache file;

the visualization description file comprises basic information of the target tree model, the model structure visualization file comprises structure information of the target tree model, and the tree model visualization cache file comprises cache information of the target tree model.

It can be understood that the model display apparatus 90 according to the embodiment of the present invention can implement the processes of the method embodiment shown in fig. 1, and achieve the same technical effects, and therefore, for avoiding repetition, the details are not repeated here.

In addition, an embodiment of the present invention further provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and capable of running on the processor, where the computer program, when executed by the processor, can implement each process of the method embodiment shown in fig. 1 and achieve the same technical effect, and is not described herein again to avoid repetition.

The embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, can implement each process of the method embodiment shown in fig. 1 and achieve the same technical effect, and is not described herein again to avoid repetition.

Computer-readable media, which include both non-transitory and non-transitory, removable and non-removable media, may implement the information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.

It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.

The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.

Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solution of the present invention or the portions contributing to the prior art may be essentially embodied in the form of a software product, which is stored in a storage medium (such as ROM/RAM, magnetic disk, optical disk) and includes instructions for enabling a service classification device (which may be a mobile phone, a computer, a server, an air conditioner, or a network device) to execute the method according to the embodiments of the present invention.

The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.

Claims (10)

1. A method of model display, comprising:

receiving model display information input by a user;

acquiring target node information in a model visualization file of a target tree model according to the model display information;

and performing model display based on the target node information.

2. The method of claim 1, wherein the target tree model comprises at least one sub-tree model, and wherein the model visualization file comprises at least one model structure visualization file, each model structure visualization file corresponding to a sub-tree model in the target tree model;

the obtaining of the target node information in the model visualization file of the target tree model according to the model display information includes:

acquiring target node information in a model structure visualization file of a target sub-tree model according to the model display information;

and the target sub-tree model is a sub-tree model in the target tree model.

3. The method of claim 2, wherein the model presentation information comprises target tree information, starting node information, and number of layers to be presented;

the obtaining of the target node information in the model structure visualization file of the target sub-tree model according to the model display information includes:

determining a model structure visualization file of a target sub-tree model corresponding to the target tree information according to the target tree information;

and selecting the target node information from the model structure visualization file of the target sub-tree model based on the starting node information and the number of layers to be displayed.

4. The method according to claim 2 or 3, wherein each content in the model structure visualization file corresponds to each node information of the target sub-tree model;

the model display based on the target node information comprises the following steps:

loading the target node information according to rows;

and performing model display based on the loaded node information.

5. The method of claim 1, further comprising:

receiving a second input of the user on the first node of the currently displayed subtree model;

responding to the second input, and judging whether the display layer where the first node is located exceeds a preset display layer;

determining a target node of a decision path of the first node on an Nth display layer under the condition that the display layer where the first node is located exceeds the preset display layer; n is greater than 0 and less than the number of layers of the preset display layer;

and removing part or all of the sibling nodes of the target node and the child nodes of the part or all of the sibling nodes from the currently displayed subtree model.

6. A model display apparatus, comprising:

the first receiving module is used for receiving model display information input by a user;

the first acquisition module is used for acquiring target node information in a model visualization file of a target tree model according to the model display information;

and the display module is used for carrying out model display based on the target node information.

7. The apparatus of claim 6, wherein the target tree model comprises at least one sub-tree model, and wherein the model visualization file comprises at least one model structure visualization file, each model structure visualization file corresponding to one sub-tree model in the target tree model;

the first obtaining module is specifically configured to:

acquiring target node information in a model structure visualization file of a target sub-tree model according to the model display information;

and the target sub-tree model is a sub-tree model in the target tree model.

8. The apparatus of claim 7, wherein the model presentation information comprises target tree information, start node information, and number of layers to be presented;

the first obtaining module comprises:

the determining unit is used for determining a model structure visualization file of a target sub-tree model corresponding to the target tree information according to the target tree information;

and the selecting unit is used for selecting the target node information from the model structure visualization file of the target sub-tree model based on the starting node information and the number of layers to be displayed.

9. The apparatus according to claim 7 or 8, wherein each content in the model structure visualization file corresponds to each node information of the target sub-tree model;

the display module comprises:

the loading unit is used for loading the target node information according to rows;

and the display unit is used for carrying out model display based on the loaded node information.

10. The apparatus of claim 6, further comprising:

the third receiving module is used for receiving a second input of the user on the first node of the currently displayed subtree model;

the second judgment module is used for responding to the second input and judging whether the display layer where the first node is located exceeds a preset display layer or not;

the determining module is used for determining a target node of a decision path of the first node on an Nth display layer under the condition that the display layer where the first node is located exceeds the preset display layer; n is greater than 0 and less than the number of layers of the preset display layer;

and the removing module is used for removing part or all of the brother nodes of the target node and the child nodes of the part or all of the brother nodes from the currently displayed subtree model.

Images