| Revision | 37139e651931737bdbc0cd56fdd463686b017ce7 (tree) |
|---|---|
| Zeit | 2014-02-11 05:34:40 |
| Autor | Lorenzo Isella <lorenzo.isella@gmai...> |
| Commiter | Lorenzo Isella |
I modified the code to run the random forest for the Kaggle loan competition.
Python-codes/loan-minimal.py (diff)| @@ -35,7 +35,10 @@ | ||
| 35 | 35 | # loss = n.genfromtxt('loss_data.dat',dtype=float) |
| 36 | 36 | |
| 37 | 37 | |
| 38 | -train = pd.read_csv('train_data_fixed.csv') | |
| 38 | +# train = pd.read_csv('train_data_fixed.csv') | |
| 39 | + | |
| 40 | + | |
| 41 | +train = n.loadtxt('train_data_fixed_no_header.dat',dtype="float") | |
| 39 | 42 | |
| 40 | 43 | #but I still use numpy for this one as it will return the right 1D array |
| 41 | 44 |
| @@ -48,7 +51,16 @@ | ||
| 48 | 51 | # I am now already cleaning and scaling the data in an R code, so I do not need |
| 49 | 52 | # to redo this now |
| 50 | 53 | |
| 51 | -# train_balance = s.mean(train, axis=0) | |
| 54 | +# train_mean = s.mean(train, axis=0) | |
| 55 | + | |
| 56 | +# train_std= s.std(train, axis=0) | |
| 57 | + | |
| 58 | +# train=train-train_mean | |
| 59 | +# train=train/train_std | |
| 60 | + | |
| 61 | +# n.savetxt("train_mean.dat", train_mean) | |
| 62 | +# n.savetxt("train_std.dat", train_std) | |
| 63 | + | |
| 52 | 64 | |
| 53 | 65 | # print "s.shape(train_balance) is, " |
| 54 | 66 |
Fork