[英]Calculate linear regression slope matrix (Same to correlation matrix) - Python/Pandas
How do I calculate slope of each columns below?
如何计算下面每列的斜率? The.corr method scans all columns and find the correlation coeffficient with each column.
.corr 方法扫描所有列并找到每列的相关系数。 I want to do the same but get slope of the datasets with each other.
我想做同样的事情,但得到数据集的斜率。 The slope line of code below does not return expected values.
下面的代码斜线不返回预期值。 I think I am not using newaxis correctly.
我想我没有正确使用 newaxis。 I need a 5x5 symmetric matrix with 1 as diagonal which means slope is for y = x.
我需要一个 1 作为对角线的 5x5 对称矩阵,这意味着斜率为 y = x。 I did it on excel picture attached with the reference being 717024 column.
我是在 excel 图片上完成的,参考为 717024 列。 I need to iterate each column and make it as reference.
我需要迭代每一列并将其作为参考。
allowableCorr = df2_norm.corr(method = 'pearson')
slope = allowableCorr * (df2_norm.std().values / df2_norm.std().values[:, np.newaxis])
df2_norm is: df2_norm 是:
count 716865 716873 716884 716943 716944
0 -0.16029615828413712 -0.07630309240006158 0.11220663712532133 -0.2726775504078691 -0.23279127015045065
1 -0.6687265363491811 -0.6135022705188075 -0.49097425130988914 -0.736020384028633 -0.705286321388766
2 0.06735205699309535 0.07948417451634422 0.09240256047258057 0.0617964313591086 0.06344003100365293
3 0.372935701728449 0.44324822316416074 0.5625073287879649 0.3199599294007491 0.3420770859108217
4 0.39439310866886124 0.45960496068147993 0.5591549439131621 0.34928093849248304 0.36951024291102974
5 -0.08007381002566456 -0.021313801077641505 0.11996141286735541 -0.15572679401876433 -0.12936514230689095
6 0.20853071107951396 0.26561990841073535 0.3661990387594055 0.15720649076873264 0.177890807311781
7 -0.0488049712326824 0.02909288268076153 0.18643283476719688 -0.1438092892727158 -0.10871022227142838
8 0.017648470149950992 0.10136455179350337 0.2722686729095633 -0.07928001803992157 -0.043102822045971705
why not为什么不
slope = np.dot(np.dot(np.diagflat(df2_norm.std().values) , allowableCorr.values), np.diagflat(1/df2_norm.std().values))
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.