Analysis

[dimaqq]

I wish there was an easy to read analysis of this data.

I've cobbled together PCA for XCode 15:

     Device       CPU       RAM       SSD  Time(sec)
0  0.382100  0.522695  0.463187  0.388886  -0.463697
1 -0.611963  0.114498  0.426237  0.477084   0.450671
2  0.279203 -0.431598 -0.371026  0.773082   0.021298
3  0.623306 -0.145030  0.401839 -0.130904   0.641752
4  0.114177  0.711597 -0.551947  0.079795   0.411803


Explained Variance Ratio:
[0.64669074 0.23646144 0.08006106 0.02745232 0.00933444]

Original variables affecting time the most:
Index(['CPU', 'RAM', 'Time(sec)', 'Device', 'SSD'], dtype='object')

device is device year; cpu is number of cores

And cross-correlation:

Cross-correlation with Time(sec):
Time(sec)    1.000000
Device       0.839492
CPU          0.725470
RAM          0.445738
SSD          0.332274
dtype: float64

Then again, my math is a bit rusty and these will change over time, so maybe someone could automate the computation?