This survey explores the advancements and challenges in using Large Language Models (LLMs) for mathematical reasoning. It delves into four key dimensions: a comprehensive exploration of various mathematical problems and their datasets, an examination of LLM-oriented techniques for solving math problems, an overview of factors affecting LLM performance in math, and a critical discussion of persistent challenges in this field. The survey highlights the importance of LLMs in automating complex tasks and their potential to enhance mathematical understanding, while also addressing the limitations and ongoing efforts to improve their capabilities. The paper provides a holistic perspective on the current state, achievements, and future directions in LLM-driven mathematical reasoning.This survey explores the advancements and challenges in using Large Language Models (LLMs) for mathematical reasoning. It delves into four key dimensions: a comprehensive exploration of various mathematical problems and their datasets, an examination of LLM-oriented techniques for solving math problems, an overview of factors affecting LLM performance in math, and a critical discussion of persistent challenges in this field. The survey highlights the importance of LLMs in automating complex tasks and their potential to enhance mathematical understanding, while also addressing the limitations and ongoing efforts to improve their capabilities. The paper provides a holistic perspective on the current state, achievements, and future directions in LLM-driven mathematical reasoning.