def __init__(self, cellmap_df, paramlist, log = False): self.cellmap_df = cellmap_df self.log = log self.cellutil = cellmaputil.cellmaputil() self.prefix_list = [] self.paramlist = paramlist for col in self.cellmap_df.columns.values: if 'source_id' in col: self.prefix_list.append(col.replace('source_id', ''))
type=config.trainingtype, logcycle=config.logcycle) training_func.starttraining() log.info('Getting trained cell map') traicellmap = training_func.givecellmap() log.info('Saving cell map in csv form') cellmapdumpfile = os.path.join(config.projectpath, 'trainedcellmap.csv') log.debug('Saving to: %s' % cellmapdumpfile) traicellmap.to_csv(cellmapdumpfile, index_label=False, index=False) log.info('Finish Saving cell map in csv form') if 'mapping' in config.processlist: if not 'cellmap_df' in locals(): log.info('Loading cellmap data frame from project file: %s' % config.projectpath) cellmaputil_func = cellmaputil.cellmaputil() cellmap_df = cellmaputil_func.loadcellmap(os.path.join(config.projectpath, 'trainedcellmap.csv')) log.info('Begin mapping source to cell map') #primary if 'dev' in config.processlist: import random _srtscalledsource_df = primerawsource_df.ix[ random.sample(primerawsource_df.index , 1000)] mapping_func = mapping2.mapping(cellmap_df, _srtscalledsource_df, config.projectpath, config.paramlist , plot=True, log=log, type=config.trainingtype) else: mapping_func = mapping2.mapping(cellmap_df, primerawsource_df, config.projectpath, config.paramlist, plot=True, log=log, type=config.trainingtype) mappedcellmap_df = mapping_func.mapsource()