diff --git a/README.md b/README.md
index 1aafff4..af00476 100644
--- a/README.md
+++ b/README.md
@@ -88,7 +88,7 @@ pip-compile
 
 Initializing a dbt project:
 ```bash
-dbt init healthcare_data
+dbt init data_processing
 ```
 
 ## Building the datasets
@@ -96,7 +96,7 @@ dbt init healthcare_data
 1. Generate the synthetic healthcare data schemas using the data dictionary:
 
 ```bash
-cd healthcare_data
+cd data_processing
 python scripts/generate_syh_dr_data_models.py ~/data/syh_dr https://www.ahrq.gov/sites/default/files/wysiwyg/data/SyH-DR-Codebook.pdf
 ```
 
@@ -109,7 +109,7 @@ dbt run --threads 8
 3. Verify that you can query the data on the command line:
 
 ```bash
-duckdb -c "SELECT * FROM '/Users/me/data/syh_dr/syhdr_commercial_inpatient_2016.parquet'"
+duckdb -c "SELECT * FROM '~/data/syh_dr/syhdr_commercial_inpatient_2016.parquet'"
 ```
 
 This should show the data:
diff --git a/data_processing/scripts/generate_syh_dr_data_models.py b/data_processing/scripts/generate_syh_dr_data_models.py
index eacef23..1c42d51 100644
--- a/data_processing/scripts/generate_syh_dr_data_models.py
+++ b/data_processing/scripts/generate_syh_dr_data_models.py
@@ -137,7 +137,9 @@ def process_csv_files(pdf_url, csv_folder):
                 print(csv_str)
             username = os.environ.get("USER")
             path_without_user = "~/" + csv_path.split(username + '/')[1]
-            select_statement = f"SELECT\n    {',\n    '.join(column_list)}\nFROM read_csv('{path_without_user}', header=True, null_padding=true{csv_str if csv_types else ''})"
+            select_statement = f"""SELECT
+    {',\n    '.join(column_list)}
+FROM read_csv('{path_without_user}', header=True, null_padding=true{csv_str if csv_types else ''})"""
             f.write(select_statement)
 
         print(f"Generated SQL model: {sql_file}")