Abstract:Machine learning (ML) has been increasingly used in a variety of domains, while solving ML programming tasks poses unique challenges due to the fundamental difference in the nature and the construct of general programming tasks, especially for developers who do not have ML backgrounds. Automatic code generation that produces a code snippet from a natural language description can be a promising technique to accelerate ML programming tasks. In recent years, although many deep learning-based neural code generation models have been proposed with high accuracy, the fact that most of them are mainly evaluated on general programming tasks calls into question their effectiveness and usefulness in ML programming tasks. In this paper, we set out to investigate the effectiveness of existing neural code generation models on ML programming tasks. For our analysis, we select six state-of-the-art neural code generation models, and evaluate their performance on four widely used ML libraries, with newly created 83K pairs of natural-language described ML programming tasks. Our empirical study reveals some good, bad, and missing aspects of neural code generation models on ML tasks, with a few major ones listed below. ( Good ) Neural code generation models perform significantly better on ML tasks than on non-ML tasks with an average difference of 10.6 points in BLEU-4 scores. ( Bad ) More than 80% of the generated code is semantically incorrect. ( Bad ) Code generation models do not have significance in improving developers’ completion time. ( Good ) The generated code can help developers write correct code by providing developers with clues for using correct APIs. ( Missing ) The observation from our user study reveals the missing aspects of code generation for ML tasks, e.g., decomposing code generation for divide-and-conquer into API sequence identification and API usage generation.

Automatic Generation of Pseudocode with Attention Seq2seq Model

From Code to Natural Language: Type-Aware Sketch-Based Seq2Seq Learning

Fine-grained Pseudo-code Generation Method via Code Feature Extraction and Transformer

A Neural Model for Generating Natural Language Summaries of Program Subroutines

code2seq: Generating Sequences from Structured Representations of Code

Scaling Automatic Extraction of Pseudocode

Neural Comment Generation for Source Code with Auxiliary Code Classification Task

A Novel Syntax-Aware Automatic Graphics Code Generation with Attention-Based Deep Neural Network

CODEP: Grammatical Seq2Seq Model for General-Purpose Code Generation.

Pseudocode to Code Based on Adaptive Global and Local Information

Automatic Source Code Summarization with Graph Attention Networks

CodeAttention: Translating Source Code to Comments by Exploiting the Code Constructs

PinNet: Pinpoint Instructive Information for Retrieval Augmented Code-to-Text Generation

Automatic Code Annotation Generation Based on Heterogeneous Graph Structure

Retrieve and Refine: Exemplar-based Neural Comment Generation

Generating Question Titles for Stack Overflow from Mined Code Snippets

The Source Code Comment Generation Based on Deep Reinforcement Learning and Hierarchical Attention

SkCoder: A Sketch-based Approach for Automatic Code Generation

The Good, the Bad, and the Missing: Neural Code Generation for Machine Learning Tasks

Compilable Neural Code Generation with Compiler Feedback

Automatic Generation of Text Descriptive Comments for Code Blocks