Spaces:
Running
on
CPU Upgrade
Running
on
CPU Upgrade
| import pandas as pd | |
| path_to_data = "./docStore/" | |
| from appStore.prep_utils import create_chunks | |
| def process_iati(): | |
| """ | |
| this will read the iati files and create the chunks | |
| """ | |
| orgas_df = pd.read_csv(f"{path_to_data}iati_files/project_orgas.csv") | |
| region_df = pd.read_csv(f"{path_to_data}iati_files/project_region.csv") | |
| sector_df = pd.read_csv(f"{path_to_data}iati_files/project_sector.csv") | |
| status_df = pd.read_csv(f"{path_to_data}iati_files/project_status.csv") | |
| texts_df = pd.read_csv(f"{path_to_data}iati_files/project_texts.csv") | |
| projects_df = pd.merge(orgas_df, region_df, on='iati_id', how='inner') | |
| projects_df = pd.merge(projects_df, sector_df, on='iati_id', how='inner') | |
| projects_df = pd.merge(projects_df, status_df, on='iati_id', how='inner') | |
| projects_df = pd.merge(projects_df, texts_df, on='iati_id', how='inner') | |
| projects_df = projects_df[projects_df.client.str.contains('bmz')].reset_index(drop=True) | |
| projects_df.drop(columns= ['orga_abbreviation', 'client', | |
| 'orga_full_name', 'country', | |
| 'country_flag', 'crs_5_code', 'crs_3_code','country_code_list', | |
| 'sgd_pred_code','crs_5_name', 'crs_3_name', 'sgd_pred_str'], inplace=True) | |
| #print(projects_df.columns) | |
| projects_df['text_size'] = projects_df.apply(lambda x: len((x['title_main'] + x['description_main']).split()), axis=1) | |
| projects_df['chunks'] = projects_df.apply(lambda x:create_chunks(x['title_main'] + x['description_main']),axis=1) | |
| projects_df = projects_df.explode(column=['chunks'], ignore_index=True) | |
| projects_df['source'] = 'IATI' | |
| projects_df.rename(columns = {'iati_id':'id','iati_orga_id':'org'}, inplace=True) | |
| return projects_df | |
| def process_giz_worldwide(): | |
| """ | |
| this will read the giz_worldwide files and create the chunks | |
| """ | |
| giz_df = pd.read_json(f'{path_to_data}giz_worldwide/data_giz_website.json') | |
| giz_df = giz_df.rename(columns={'content':'project_description'}) | |
| giz_df['text_size'] = giz_df.apply(lambda x: len((x['project_name'] + x['project_description']).split()), axis=1) | |
| giz_df['chunks'] = giz_df.apply(lambda x:create_chunks(x['project_name'] + x['project_description']),axis=1) | |
| print("initial df length:",len(giz_df)) | |
| giz_df = giz_df.explode(column=['chunks'], ignore_index=True) | |
| print("new df length:",len(giz_df)) | |
| print(giz_df.columns) | |
| #giz_df.drop(columns = ['filename', 'url', 'name', 'mail', | |
| # 'language', 'start_year', 'end_year','poli_trager'], inplace=True) | |
| giz_df['source'] = 'GIZ_WORLDWIDE' | |
| return giz_df | |
| def remove_duplicates(results_list): | |
| """ | |
| Return a new list of results with duplicates removed, | |
| based on 'url' in metadata. | |
| """ | |
| unique_results = [] | |
| seen_urls = set() | |
| for r in results_list: | |
| # Safely get the URL from metadata | |
| url = r.payload['metadata'].get('url', None) | |
| if url not in seen_urls: | |
| seen_urls.add(url) | |
| unique_results.append(r) | |
| return unique_results | |