init

.gitattributes

@@ -0,0 +1,2 @@
+test.csv filter=lfs diff=lfs merge=lfs -text
+train.csv filter=lfs diff=lfs merge=lfs -text

process.py

@@ -0,0 +1,21 @@
+#%% 
+import h5py
+import pandas as pd
+import numpy as np
+
+#%%
+with h5py.File("usps.h5", 'r') as hf:
+    train = hf.get('train')
+    X_tr = train.get('data')[:]
+    y_tr = train.get('target')[:]
+    test = hf.get('test')
+    X_te = test.get('data')[:]
+    y_te = test.get('target')[:]
+
+# %%
+X_training = np.concatenate((X_tr, y_tr.reshape(-1, 1)), axis=1)
+X_test = np.concatenate((X_te, y_te.reshape(-1, 1)), axis=1)
+
+# %%
+pd.DataFrame(X_training).to_csv("train.csv", index=False, header=False)
+pd.DataFrame(X_test).to_csv("test.csv", index=False, header=False)

test.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:229900de252b803e4cf1dd75c70d6495bf82e7393cb1c250bd2c07abfd8d4a6d
+size 4509824

train.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:39097e4dd1ab3b3079ca5286d7ea210bdddd9970f8d04932cb3f9a2032bdaa0b
+size 16341487