From 905a2e245990d820a293dd933cc71f0db637bc63 Mon Sep 17 00:00:00 2001
From: JunhaoQiu <56094690+qchiujunhao@users.noreply.github.com>
Date: Tue, 2 Apr 2024 17:06:57 -0400
Subject: [PATCH] Auto generate a config from a dataset

---
 tools/ludwig_autogencofig.py  | 25 +++++++++++++++++++++
 tools/ludwig_autogencofig.xml | 41 +++++++++++++++++++++++++++++++++++
 2 files changed, 66 insertions(+)
 create mode 100644 tools/ludwig_autogencofig.py
 create mode 100644 tools/ludwig_autogencofig.xml

diff --git a/tools/ludwig_autogencofig.py b/tools/ludwig_autogencofig.py
new file mode 100644
index 0000000..8f9d7f7
--- /dev/null
+++ b/tools/ludwig_autogencofig.py
@@ -0,0 +1,25 @@
+import argparse
+
+import logging
+
+from ludwig import automl
+from ludwig.utils import defaults
+
+logging.basicConfig(level=logging.DEBUG)
+LOG = logging.getLogger(__name__)
+
+def main():
+    parser = argparse.ArgumentParser(description='Render a Ludwig config')
+    parser.add_argument('--dataset', type=str, help='Path to the dataset file', required=True)
+    parser.add_argument('--output_feature', type=str, help='Name for the output feature', required=True)
+    parser.add_argument('--output', type=str, help='Path for the output file', required=True)
+    args = parser.parse_args()
+
+    
+    args_init = ["--dataset", args.dataset, "--target", args.output_feature, "--output", args.output]
+    automl.cli_init_config(args_init)
+    args_render = ["--config", args.output, "--output", args.output]
+    defaults.cli_render_config(args_render)
+
+if __name__ == "__main__":
+    main()
\ No newline at end of file
diff --git a/tools/ludwig_autogencofig.xml b/tools/ludwig_autogencofig.xml
new file mode 100644
index 0000000..83da665
--- /dev/null
+++ b/tools/ludwig_autogencofig.xml
@@ -0,0 +1,41 @@
+<tool id="ludwig_config_generator" name="Ludwig Config Generator" version="@VERSION@" profile="@PROFILE@">
+    <description>Auto-generate a config file from a dataset for Ludwig training.</description>
+    <macros>
+        <import>ludwig_macros.xml</import>
+    </macros>
+    <expand macro="python_requirements" />
+    <expand macro="macro_stdio" />
+    <version_command>echo "@VERSION@"</version_command>
+    <command>
+        <![CDATA[
+            #if $input_file
+            ln -sf '$input_file' "`pwd`/${input_file.element_identifier}";
+            #end if
+            python '$__tool_directory__/ludwig_autogencofig.py' 
+                #if $input_file   
+                --dataset "`pwd`/${input_file.element_identifier}"
+                #end if
+                --output_feature '$output_feature'
+                --output '$output'
+        ]]>
+    </command>
+    <inputs>
+        <param name="input_file" type="data" label="Select CSV File" format="tabular,csv,h5,json,txt" help="Choose a dataset file for processing">
+    
+        </param>
+    
+    <!-- <param name="output_feature" type="" data_ref="input_file" multiple="True" numerical="False" label="Output Feature"> -->
+    <param argument="output_feature" type="text" value="" label="Output Feature" />
+    <param name="config" type="text" area="true" />
+    </inputs>
+
+    <outputs>
+        <data name="output" format="yaml" label="${tool.name}" />
+    </outputs>
+
+    <tests>
+        <test>
+            <!-- TO DO -->
+        </test>
+    </tests>
+</tool>