主成分分析减少了行数 [英] Principal Component Analysis decreases the number of rows

问题描述

当我在Azure ML Studio中运行Principal Component Analysis时，输出具有新组件，但观察次数减少，任何人都可以解释为什么以及如何解决此问题？

When i run Principal Component Analysis in Azure ML Studio, the output has the new components but the number of observation diminishes, can anyone explain why and how to solve this problem ?

推荐答案

你好，

如果是PCA，如果源数据集有八列，你输入3作为要减少的维度，则会返回三个新列八个选定列的信息。列名为Col1，Col2和Col3。这些列的
不直接映射到源列;相反，列包含原始列描述的特征空间的近似值。

In case of PCA if the source dataset has eight columns and you type 3 as the dimensions to reduce to, three new columns are returned that capture the information of the eight selected columns. The columns are named Col1, Col2, and Col3. These columns do not map directly to the source columns; instead, the columns contain an approximation of the feature space described by the original columns.

如果您看到的行少于您在可视化输出中所期望的行，则可能是显示只显示100行，您可以通过转换为csv来下载数据集以检查结果集。 

In case if you are seeing rows less than what you expected in the visualize output it could be that the display shows only 100 rows, you can download the dataset by converting to csv to check the result-set. 

从visualize的输出结果中看到的数据显示1000行但是它只能在此窗口中显示最多100行。

Data as seen in the result of my output from visualize shows 1000 rows but it can display upto 100 rows only in this window.

下载为CSV文件，它将显示所有行。

Download as a CSV file and it will show all the rows.

以下是csv文件中的行数。

Here is the count of the rows in the csv file.

我希望这会有所帮助!!

I hope this helps!!

这篇关于主成分分析减少了行数的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！