《QC工具之-散布图雷达图汇总课件.ppt》由会员分享，可在线阅读，更多相关《QC工具之-散布图雷达图汇总课件.ppt（27页珍藏版）》请在三一办公上搜索。
1、工具之,-,散布图、雷达图,QC,?#?,散布图,-,定义,散布图定义,为研究两个变量之间的相关性，而搜集成对两组,数据，在坐标上用,点,来表示出两个特性值之间相关情,形的图形，称之为散布图。,.,.,.,.,.,.,.,.,.,?#?,散布图,-,作用,散布图作用,产,量,与,温,度,_,1,的,散,点,图,1,、收集原因地数据与结果进行比较，可,发现原因与结果的关系；,2,、在散布图内将原因和结果的数据填,200,175,150,产,量,入，绘出散布图对结果可一目了然。,3,、由散布图可清楚的了解两组数据间的,关系，可以判断是否有关联。,125,100,75,50,0,5,10,温,度,_
2、,1,15,20,25,?#?,散布图,-,做法,散布图做法,某产品的烧熔温度及,硬度,之间是否存在有相关,性，今收集,30,组数据，请予以分析。,步骤,1,：收集,30,组以上的相对数据,整理到数据,表上。,数据不能太少，否则易生误判,?#?,散布图,-,做法,X,Y,X,Y,X,Y,NO.,燒溶溫度,硬度,NO.,燒溶溫度,硬度,NO.,燒溶溫度,硬度,1,2,3,4,5,6,7,8,9,10,810,890,850,840,890,870,860,810,820,830,47,56,48,45,54,59,50,51,42,53,11,12,13,14,15,16,17,18,19,20
3、,840,870,830,830,820,820,830,850,870,820,52,56,51,45,46,48,55,55,49,44,21,22,23,24,25,26,27,28,29,30,810,850,880,880,840,880,830,860,840,830,44,53,54,57,50,54,53,51,50,49,步骤,2,：找出数据,X,、,Y,的最大值及最小值。,步骤,3,：画出纵轴与横轴。,两种数据，一方为原因、另一方为结果时，横轴取原因，纵轴取结果。,?#?,散布图,-,做法,步骤,4,：将各组成对数据标记在座标上；,横轴与纵轴的数据交会处点上“,”；,两组数
4、据重复在同一点上时，划双重圆记号；,三组数据重复在同一点上时，划三重圆记号。,有层别项目时，一方用,?,来区分，另一方用来分类。看出错误数据，,并将其排除在外。,步骤：记下必要事项：数据、采取時间、目的、产品名,、工程名、绘图者、绘制日期等，,判断（分析研,究点子云的分布状况，确定相关关系的类型）,。,?#?,散布图,-,做法,N=30,时间段：,99.12.20-99.12.30,硬,度,Y,.,.,.,.,.,.,.,.,.,.,.,.,.,.,.,.,.,.,.,产品名：,SI-083,单位：压延课,绘制：,日期：,99.12.31,烧熔溫度,X,?#?,散布图,-,相关性判断,散布图的
5、相关性判断,1.,对照典型图例判断法,2.,象限判断法,3.,相关系数判断法,?#?,散布图,-6,类型,Y,Y,强正相关,强负相关,0,Y,不相关,0,弱正相关,Y,0,X,0,Y,弱负相关,X,X,0,Y,X,非直线相关,（曲线）,X,0,X,?#?,散布图,-,实例,实例钢的淬火温度与硬度的相关关系判断,序号,1,2,3,4,5,6,7,8,9,0,11,12,13,14,15,淬火温度（,C,0,）,硬度（,HRC,）,x,Y,810,890,850,840,850,890,870,860,810,820,840,870,830,830,820,47,56,48,45,54,59,50
6、,51,42,53,52,53,51,45,46,序号,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,淬火温度（,C,0,）,硬度（,HRC,）,x,Y,820,860,870,830,820,810,850,880,880,840,880,830,860,860,840,48,55,55,49,44,44,53,54,57,50,54,46,52,50,49,?#?,散布图,-,实例,60,58,1,、,对,照,典,型,图,例,判,断,法,56,硬,度,（,H,R,C,）,54,52,50,48,46,44,42,810,820,830,840,8
7、50,860,870,880,890,?#?,淬火温度（,）,散布图,-,实例,2,、,象限判断法,?,象限判断法又叫中值判断法、符号检定判断法。,使用此法的步骤如下：,在散布图上画一条与,Y,轴平行的中值线,f,，使,f,线的左、右两边的,点子数大致相等；,在散布图上画一条与,X,轴平行的中值线,g,，使,g,线的上、下两边的,点子数大致相等；,f,、,g,两条线把散布图分成,4,个象限区域,I,、,II,、,III,、,IV,。分别统计,落入各象限区域内的点子数；,分别计算对角象限区内的点子数；,判断规则；,若,n,I,n,III,n,II,n,IV,，则判为正相关,若,n,I,n,III
8、,n,II,n,IV,，则判为负相关,?#?,散布图,-,实例,60,f,58,硬,度,（,H,R,C,）,56,54,52,50,48,g,46,44,42,810,820,830,840,850,860,870,880,890,?#?,淬火温度（,）,散布图,-,实例,3,、,相关系数判断法,相关系数判断法的应用步骤：,1.,2.,简化,X,、,Y,数据。,计算,X,2,，,Y,2,，,X Y,、（,X,Y,）和（,X,Y,）,2,。,3.,4.,计算,X,、,Y,、,X Y,、,X,2,、,Y,2,、,（,X,Y,）和,计算,L,X X,、,L,Y Y,、,L,X Y,。,2,（,X,）
9、,L,X X,=,X,2,N,2,（,Y,）,L,Y Y,=,Y,2,N,（,X,Y,）,2,。,L,X Y,=,X Y,（,X,）（,Y,）,N,?#?,散布图,-,实例,5.,计算相关数据（,）。,L,X Y,L,X X,L,Y Y,6.,查出临界相关数据（,）。,7.,判断。判断规则：,若,，则,X,与,Y,相关,若,，则,X,与,Y,不相关,?#?,散布图,-,实例,N,O,1,2,3,4,5,6,7,8,9,10,11,12,13,14,X,1,9,5,4,5,9,7,6,1,2,4,7,3,3,Y,7,16,8,5,14,19,10,11,2,13,12,13,11,5,X,2,1
10、,81,25,16,25,81,49,36,1,4,16,49,9,9,Y,2,49,256,64,25,196,361,100,121,4,169,144,169,121,25,X Y,7,144,40,20,70,171,70,66,2,26,48,91,33,15,X,Y,8,25,13,9,19,28,17,17,3,15,16,20,14,8,（,X,Y,）,2,64,625,169,81,361,784,289,289,9,225,256,400,196,64,15,2,6,4,36,12,8,64,?#?,散布图,-,实例,N,O,16,17,X,2,6,Y,8,15,X,2,4
11、,36,Y,2,64,225,X Y,16,90,X,Y,10,21,（,X,Y,）,2,100,441,18,19,20,21,7,3,2,1,15,9,4,4,49,9,4,1,225,81,16,16,105,27,8,4,22,12,6,5,484,144,36,25,22,23,24,25,5,8,8,4,13,14,17,10,25,64,64,16,169,196,289,100,65,112,136,40,18,22,25,14,324,484,625,196,26,27,28,29,8,3,6,6,14,6,12,10,64,9,36,36,196,36,144,100,11
12、2,18,72,60,22,9,18,16,484,81,324,256,30,合计,4,141,9,312,16,839,81,3778,36,1716,13,453,169,8049,?#?,散布图,-,实例,注：,表中,X,值是（,X,800,）,1/10,的简化值；,Y,值是（,Y,40,）,1,的简化值。,表中,X,Y,、（,X,Y,）,2,栏是校对栏，以免,X,、,Y,、,X,2,、,Y,2,、,X,Y,各,栏计算错误，导致相关性结论错误。校核公式是：,（,X,Y,）,X,Y,（,X,Y,）,2,X,2,2,（,X,Y,）,Y,2,4.,计算,L,X X,、,L,Y Y,、,L,X
13、 Y,。,2,2,（,X,）,（,141,）,839,L,X X,=,X,2,176.3,N,30,2,2,（,Y,）,（,312,）,3778,L,Y Y,=,Y,2,533.2,N,30,L,X Y,=,X Y,（,X,）（,Y,）,1716,141,312,30,249.6,?#?,散布图,-,实例,5.,计算相关系数（,）。,L,X Y,L,X X,L,Y Y,249.6,176.3,533.2,0.814,6.,查出临界相关数据（,）。,根据,N,2,和显著性水平,查表求得,0.361,（,0.05,）,7.,判断。判断规则：,0.814,0.361,，所以钢的硬度与淬火温度呈强正相
14、关。,以上三种判断方法对同一实例进行分析判断的结论是一致的。,后附相关系数检查表,?#?,散布图,-,相关系数,r,取值,r=1,现象,n,个点全在拟合直线上,意义,7,6,图示,5,完全正相关,Y,4,3,1,2,3,X,4,5,6,5,4,0r1,n,个点随机分散在拟合直线两,边，,正相关，越大相关性,越强,3,Y,2,1,0,1,2,3,X,4,5,4,3,r=0,n,个点分布无一点规律，或某,种弯曲趋势,n,个点随机分散在拟合直线两,边，,2,无线性相关性,Y,1,0,1,2,3,X,4,5,6,5,4,-1r0,负相关，越小相关性,越强,3,Y,2,1,0,1,2,3,X,4,5,7
15、,6,5,r=-1,n,个点全在拟合直线上,完全负相关,Y,4,3,1,2,3,X,4,5,?#?,散布图,-,相关系数检查表,N,2,1,2,3,0.05,0.997,0.950,0.878,0.01,1.000,0.990,0.959,N,2,11,12,13,0.05,0.553,0.532,0.514,0.01,0.684,0.661,0.641,4,5,6,7,0.811,0.754,0.707,0.666,0.917,0.874,0.834,0.798,14,15,16,17,0.497,0.482,0.468,0.456,0.623,0.606,0.590,0.575,8,9,1
16、0,0.632,0.602,0.675,0.765,0.735,0.708,18,19,20,0.444,0.433,0.423,0.561,0.549,0.537,?#?,散布图,-,相关系数检查表,N,2,21,0.05,0.413,0.01,0.526,N,2,35,0.05,0.325,0.01,0.418,22,23,24,25,26,27,28,29,30,0.404,0.396,0.388,0.381,0.374,0.367,0.361,0.355,0.349,0.515,0.505,0.496,0.487,0.478,0.470,0.463,0.456,0.449,40,45,
17、50,60,70,80,90,100,200,0.304,0.288,0.273,0.250,0.232,0.217,0.205,0.195,0.138,0.393,0.372,0.354,0.325,0.302,0.283,0.267,0.254,0.181,?#?,散布图,-,判读注意事项,散布图判读注意事项,1,）注意有无异常点,1.,2,）是否有层别必要,3,）是否为假相关,4,）勿依据技术、经验作直觉的判断,5,）数据太少，容易发生误判,显著性水平,在统计,学中叫做犯第一类错误,的大小，第一类错误就,是原假设是对的，但是,被拒绝的概率，我们一,般把这个显著性水平,定,为,0.05,。,?#?,散点图,定义：,散点图是研究成对数据出现的两组数据,之间相关关系的简单图示。,案例：,某团队在分析产品加工温度与产量之间的关系,时收集了一批数据，绘制散点图如右所示,产,量,产,量,与,温,度,_,1,的,散,点,图,200,175,150,125,100,75,作用：,主要是研究两个变量之间存在何种关系,50,0,5,10,温,度,_,1,15,20,25,?#?,雷达图,GO,?#?,?#?,此课件下载可自行编辑修改，仅供参考！,感谢您的支持，我们努力做得更好！谢谢,?#?,
