3 files changed, 149 insertions, 0 deletions
diff --git a/challenge-307/conor-hoekstra/apl/ch-1.apl b/challenge-307/conor-hoekstra/apl/ch-1.apl
new file mode 100644
index 0000000000..53a183db51
--- /dev/null
+++ b/challenge-307/conor-hoekstra/apl/ch-1.apl
@@ -0,0 +1 @@
+CheckOrder ← 1-⍨⍸(⊂⍤⍋⍛⌷)⍛≠
+\ No newline at end of file
diff --git a/challenge-307/conor-hoekstra/cuda/ch-1.cu b/challenge-307/conor-hoekstra/cuda/ch-1.cu
new file mode 100644
index 0000000000..ecab913a08
--- /dev/null
+++ b/challenge-307/conor-hoekstra/cuda/ch-1.cu
@@ -0,0 +1,5 @@
+#import "parrot.h"
+
+auto check_order(auto ints) { //
+  return ints.sort().neq(ints).where() - 1;
+}
+\ No newline at end of file
diff --git a/challenge-307/conor-hoekstra/python/ch-1.py b/challenge-307/conor-hoekstra/python/ch-1.py
new file mode 100644
index 0000000000..de99d0885e
--- /dev/null
+++ b/challenge-307/conor-hoekstra/python/ch-1.py
@@ -0,0 +1,143 @@
+import numpy as np
+import jax
+import jax.numpy as jnp
+import tensorflow as tf
+import torch
+
+
+def check_order(lst):
+    return [c for c, (a, b) in enumerate(zip(lst, sorted(lst))) if a != b]
+
+
+# NumPy Solution
+def check_order_np(lst):
+    l = np.array(lst)
+    return np.where(l != sorted(l))[0]
+
+
+def check_order_jax(lst):
+    l = jnp.array(lst)
+    return jnp.where(l != jnp.sort(l))[0]
+
+
+# @jax.jit
+# def check_order_jax_jit(lst):
+#     l = jnp.array(lst)
+#     return jnp.where(l != jnp.sort(l))[0]
+
+# jax.errors.ConcretizationTypeError: Abstract tracer value encountered
+#  where concrete value is expected: traced array with shape int32[]
+# The size argument of jnp.nonzero must be statically specified to use
+# jnp.nonzero within JAX transformations.
+
+
+# TensorFlow Solution (CPU)
+def check_order_tf(lst):
+    l = tf.constant(lst)
+    return tf.where(l != tf.sort(l))[:, 0].numpy()
+
+
+# TensorFlow Solution (CUDA/GPU)
+def check_order_tf_cuda(lst):
+    with tf.device("/GPU:0"):
+        l = tf.constant(lst)
+        return tf.where(l != tf.sort(l))[:, 0].numpy()
+
+
+# TensorFlow Function Solution (CPU)
+@tf.function
+def check_order_tf_fn(l):
+    return tf.where(l != tf.sort(l))[:, 0]
+
+
+# TensorFlow Function Solution (CUDA/GPU)
+@tf.function
+def check_order_tf_fn_cuda_impl(l):
+    return tf.where(l != tf.sort(l))[:, 0]
+
+
+def check_order_tf_fn_cuda(lst):
+    with tf.device("/GPU:0"):
+        l = tf.constant(lst)
+        return check_order_tf_fn_cuda_impl(l).numpy()
+
+
+# PyTorch Solution (CPU)
+def check_order_torch(lst):
+    l = torch.tensor(lst)
+    return torch.where(l != torch.sort(l)[0])[0].numpy()
+
+
+# PyTorch Solution (CUDA/GPU)
+def check_order_torch_cuda(lst):
+    l = torch.tensor(lst).cuda()
+    return torch.where(l != torch.sort(l)[0])[0].cpu().numpy()
+
+
+# PyTorch JIT Compile Solution (CPU)
+@torch.jit.script
+def check_order_torch_jit_impl(l):
+    return torch.where(l != torch.sort(l)[0])[0]
+
+
+def check_order_torch_jit(lst):
+    l = torch.tensor(lst)
+    return check_order_torch_jit_impl(l).numpy()
+
+
+# PyTorch JIT Compile Solution (CUDA/GPU)
+@torch.jit.script
+def check_order_torch_jit_cuda_impl(l):
+    return torch.where(l != torch.sort(l)[0])[0]
+
+
+def check_order_torch_jit_cuda(lst):
+    l = torch.tensor(lst).cuda()
+    return check_order_torch_jit_cuda_impl(l).cpu().numpy()
+
+
+# Tests
+print(check_order([5, 2, 4, 3, 1]))  # [0, 2, 3, 4]
+print(check_order([1, 2, 1, 1, 3]))  # [1, 3]
+print(check_order([3, 1, 3, 2, 3]))  # [0, 1, 3]
+print(check_order_np([5, 2, 4, 3, 1]))  # [0, 2, 3, 4]
+print(check_order_np([1, 2, 1, 1, 3]))  # [1, 3]
+print(check_order_np([3, 1, 3, 2, 3]))  # [0, 1, 3]
+print(check_order_jax([5, 2, 4, 3, 1]))  # [0, 2, 3, 4]
+print(check_order_jax([1, 2, 1, 1, 3]))  # [1, 3]
+print(check_order_jax([3, 1, 3, 2, 3]))  # [0, 1, 3]
+print(check_order_tf([5, 2, 4, 3, 1]))  # [0, 2, 3, 4]
+print(check_order_tf([1, 2, 1, 1, 3]))  # [1, 3]
+print(check_order_tf([3, 1, 3, 2, 3]))  # [0, 1, 3]
+print(check_order_tf_fn(tf.constant([5, 2, 4, 3, 1])).numpy())  # [0, 2, 3, 4]
+print(check_order_tf_fn(tf.constant([1, 2, 1, 1, 3])).numpy())  # [1, 3]
+print(check_order_tf_fn(tf.constant([3, 1, 3, 2, 3])).numpy())  # [0, 1, 3]
+print(check_order_torch([5, 2, 4, 3, 1]))  # [0, 2, 3, 4]
+print(check_order_torch([1, 2, 1, 1, 3]))  # [1, 3]
+print(check_order_torch([3, 1, 3, 2, 3]))  # [0, 1, 3]
+print(check_order_torch_jit([5, 2, 4, 3, 1]))  # [0, 2, 3, 4]
+print(check_order_torch_jit([1, 2, 1, 1, 3]))  # [1, 3]
+print(check_order_torch_jit([3, 1, 3, 2, 3]))  # [0, 1, 3]
+
+# CUDA/GPU Tests (requires CUDA-capable GPU)
+if tf.config.list_physical_devices("GPU"):
+    print("TensorFlow GPU available")
+    print(check_order_tf_cuda([5, 2, 4, 3, 1]))  # [0, 2, 3, 4]
+    print(check_order_tf_cuda([1, 2, 1, 1, 3]))  # [1, 3]
+    print(check_order_tf_cuda([3, 1, 3, 2, 3]))  # [0, 1, 3]
+    print(check_order_tf_fn_cuda([5, 2, 4, 3, 1]))  # [0, 2, 3, 4]
+    print(check_order_tf_fn_cuda([1, 2, 1, 1, 3]))  # [1, 3]
+    print(check_order_tf_fn_cuda([3, 1, 3, 2, 3]))  # [0, 1, 3]
+else:
+    print("TensorFlow GPU not available")
+
+if torch.cuda.is_available():
+    print("PyTorch CUDA available")
+    print(check_order_torch_cuda([5, 2, 4, 3, 1]))  # [0, 2, 3, 4]
+    print(check_order_torch_cuda([1, 2, 1, 1, 3]))  # [1, 3]
+    print(check_order_torch_cuda([3, 1, 3, 2, 3]))  # [0, 1, 3]
+    print(check_order_torch_jit_cuda([5, 2, 4, 3, 1]))  # [0, 2, 3, 4]
+    print(check_order_torch_jit_cuda([1, 2, 1, 1, 3]))  # [1, 3]
+    print(check_order_torch_jit_cuda([3, 1, 3, 2, 3]))  # [0, 1, 3]
+else:
+    print("PyTorch CUDA not available")