From bc36daeb8591cd84c59db5863a573a399295ea79 Mon Sep 17 00:00:00 2001
From: shivpalSW <shivpal.wadkar2050@gmail.com>
Date: Thu, 13 Jul 2023 17:37:25 +0530
Subject: [PATCH] stage -1-dvc updated

---
 artifacts/prepared/.gitignore |  2 ++
 dvc.lock                      | 31 +++++++++++++++++++++++++++++++
 dvc.yaml                      | 22 +++++++++++++++-------
 3 files changed, 48 insertions(+), 7 deletions(-)
 create mode 100644 artifacts/prepared/.gitignore
 create mode 100644 dvc.lock

diff --git a/artifacts/prepared/.gitignore b/artifacts/prepared/.gitignore
new file mode 100644
index 0000000..4478cdd
--- /dev/null
+++ b/artifacts/prepared/.gitignore
@@ -0,0 +1,2 @@
+/train.tsv
+/test.tsv
diff --git a/dvc.lock b/dvc.lock
new file mode 100644
index 0000000..0d46e9f
--- /dev/null
+++ b/dvc.lock
@@ -0,0 +1,31 @@
+schema: '2.0'
+stages:
+  prepare_data:
+    cmd: python src/stage_01_prepare.py --config=configs/config.yaml --params=params.yaml
+    deps:
+    - path: configs/config.yaml
+      md5: ef78262a4a0652b7f26da188dec74419
+      size: 230
+    - path: data/data.xml
+      md5: a304afb96060aad90176268345e10355
+      size: 37891850
+    - path: src/stage_01_prepare.py
+      md5: 47a1121c82f1e79ff3004abb9c4d5d4d
+      size: 2096
+    - path: src/utils/common.py
+      md5: 48b676b0d599169fc5e054b0829a5a4f
+      size: 519
+    - path: src/utils/data_mgmt.py
+      md5: c985e579010a81c5462ebf8184649fe4
+      size: 871
+    params:
+      params.yaml:
+        prepare.seed: 2021
+        prepare.split: 0.2
+    outs:
+    - path: artifacts/prepared/test.tsv
+      md5: 60c757f3b30604473de7fa775dfc00d9
+      size: 4899778
+    - path: artifacts/prepared/train.tsv
+      md5: 553034c4cf40efc63c99c19fe98610d0
+      size: 18986541
diff --git a/dvc.yaml b/dvc.yaml
index ad80635..d1e53f7 100644
--- a/dvc.yaml
+++ b/dvc.yaml
@@ -1,8 +1,16 @@
 # add stages here 
-# stages:
-#   stage_name:
-#     cmd: python src/stage_01______.py
-#     deps:
-#     - src/stage_01______.py
-#     outs:
-#     - output_data_of_file_path
\ No newline at end of file
+stages:
+  prepare_data:
+      cmd: python src/stage_01_prepare.py --config=configs/config.yaml --params=params.yaml
+      deps:
+        - src/stage_01_prepare.py
+        - data/data.xml
+        - src/utils/common.py
+        - src/utils/data_mgmt.py
+        - configs/config.yaml
+      params:
+        - prepare.seed
+        - prepare.split
+      outs:
+        - artifacts/prepared/train.tsv
+        - artifacts/prepared/test.tsv
\ No newline at end of file