了解成都市老年人的心理健康现状，分析影响老年人心理健康的主要因素。

2017年7月选取成都市温江区、新都区、龙泉驿区的居家老年人和养老机构老年人作为调查对象，并针对性设计调查问卷，发放问卷650份。先对调查问卷中的基本资料进行预处理并赋值，再使用k-means聚类分析算法对老年人心理健康影响因素进行分析。

问卷回收650份，回收率100%，其中有效问卷636份，问卷有效率为97.8%。最终聚类结果显示第一类案例数为117例，第二类案例数为255例，第三类案例数为264例。其中第一类案例变量属性特征主要为年龄偏大、收入偏低、生活满意度低、时常感觉人或事不顺眼、部分无配偶、儿女关心度低、常感孤独；第二类案例变量属性特征主要为年龄较大、收入较高、生活满意度高、看待生老病死乐观、儿女关心度高、生活担忧少、兴趣爱好多；第三类案例变量属性特征主要为年龄不大、收入一般、生活满意度较高、儿女关心度高。

年龄、有无配偶、收入状况及生活满意度等因素是影响老年人心理健康状况的主要因素，采用聚类分析可以获得分类结果，可为老年人的健康服务提供指导。

To study the status of mental health and its impact factors of aging population in Chengdu.

In July 2017, the elders in home and elders in the elderly districts of Wenjiang, Xindu and Longquanyi District of Chengdu were selected as the survey subjects. The questionnaire was designed specifically, and 650 questionnaires were issued. First, the basic information in the questionnaire is preprocessed and assigned, then the k-means clustering analysis algorithm is used to analyze the influencing factors of mental health in the elderly.

All 650 questionnaires were successfully collected, the recovery rate of 100%, of which 636 were valid questionnaires and the efficiency was 97.8%. The final clustering results showed that the number of the first class cases were 117 cases, the second class cases were 255 cases and the third class cases were 264 cases. The first class cases variables attribute characteristics were mainly older, low income, low life satisfaction, often feel people or things uncomfortable, some without spouses, children with low concern, and often feel lonely; the second class cases variables attribute characteristics were mainly older, higher income, high life satisfaction, optimistic view of old and sick, children with high concern, less worry about life, hobbies and more; the third class cases variables were mainly characterized by age, income, higher live satisfaction, higher degree of concern for children.

The main influencing factors to mental health of the elderly were age, marital status, income status and the degree of life satisfaction.Clustering analysis can be used to obtain classification results, which can guide the elderly's health services.