example/kaggle-ndsb1/submission_dsb.py

from __future__ import print_function
import pandas as pd
import os
import time as time

## Receives an array with probabilities for each class (columns) X images in test set (as listed in test.lst) and formats in Kaggle submission format, saves and compresses in submission_path

def gen_sub(predictions,test_lst_path="test.lst",submission_path="submission.csv"):

    ## append time to avoid overwriting previous submissions
    ## submission_path=time.strftime("%Y%m%d%H%M%S_")+submission_path

    ### Make submission
    ## check sampleSubmission.csv from kaggle website to view submission format
    header = "acantharia_protist_big_center,acantharia_protist_halo,acantharia_protist,amphipods,appendicularian_fritillaridae,appendicularian_s_shape,appendicularian_slight_curve,appendicularian_straight,artifacts_edge,artifacts,chaetognath_non_sagitta,chaetognath_other,chaetognath_sagitta,chordate_type1,copepod_calanoid_eggs,copepod_calanoid_eucalanus,copepod_calanoid_flatheads,copepod_calanoid_frillyAntennae,copepod_calanoid_large_side_antennatucked,copepod_calanoid_large,copepod_calanoid_octomoms,copepod_calanoid_small_longantennae,copepod_calanoid,copepod_cyclopoid_copilia,copepod_cyclopoid_oithona_eggs,copepod_cyclopoid_oithona,copepod_other,crustacean_other,ctenophore_cestid,ctenophore_cydippid_no_tentacles,ctenophore_cydippid_tentacles,ctenophore_lobate,decapods,detritus_blob,detritus_filamentous,detritus_other,diatom_chain_string,diatom_chain_tube,echinoderm_larva_pluteus_brittlestar,echinoderm_larva_pluteus_early,echinoderm_larva_pluteus_typeC,echinoderm_larva_pluteus_urchin,echinoderm_larva_seastar_bipinnaria,echinoderm_larva_seastar_brachiolaria,echinoderm_seacucumber_auricularia_larva,echinopluteus,ephyra,euphausiids_young,euphausiids,fecal_pellet,fish_larvae_deep_body,fish_larvae_leptocephali,fish_larvae_medium_body,fish_larvae_myctophids,fish_larvae_thin_body,fish_larvae_very_thin_body,heteropod,hydromedusae_aglaura,hydromedusae_bell_and_tentacles,hydromedusae_h15,hydromedusae_haliscera_small_sideview,hydromedusae_haliscera,hydromedusae_liriope,hydromedusae_narco_dark,hydromedusae_narco_young,hydromedusae_narcomedusae,hydromedusae_other,hydromedusae_partial_dark,hydromedusae_shapeA_sideview_small,hydromedusae_shapeA,hydromedusae_shapeB,hydromedusae_sideview_big,hydromedusae_solmaris,hydromedusae_solmundella,hydromedusae_typeD_bell_and_tentacles,hydromedusae_typeD,hydromedusae_typeE,hydromedusae_typeF,invertebrate_larvae_other_A,invertebrate_larvae_other_B,jellies_tentacles,polychaete,protist_dark_center,protist_fuzzy_olive,protist_noctiluca,protist_other,protist_star,pteropod_butterfly,pteropod_theco_dev_seq,pteropod_triangle,radiolarian_chain,radiolarian_colony,shrimp_caridean,shrimp_sergestidae,shrimp_zoea,shrimp-like_other,siphonophore_calycophoran_abylidae,siphonophore_calycophoran_rocketship_adult,siphonophore_calycophoran_rocketship_young,siphonophore_calycophoran_sphaeronectes_stem,siphonophore_calycophoran_sphaeronectes_young,siphonophore_calycophoran_sphaeronectes,siphonophore_other_parts,siphonophore_partial,siphonophore_physonect_young,siphonophore_physonect,stomatopod,tornaria_acorn_worm_larvae,trichodesmium_bowtie,trichodesmium_multiple,trichodesmium_puff,trichodesmium_tuft,trochophore_larvae,tunicate_doliolid_nurse,tunicate_doliolid,tunicate_partial,tunicate_salp_chains,tunicate_salp,unknown_blobs_and_smudges,unknown_sticks,unknown_unclassified".split(',')

        
    # read first line to know the number of columns and column to use
    img_lst = pd.read_csv(test_lst_path,sep="/",header=None, nrows=1) 
    columns = img_lst.columns.tolist() # get the columns
    cols_to_use = columns[len(columns)-1] # drop the last one
    cols_to_use= map(int, str(cols_to_use)) ## convert scalar to list


    img_lst= pd.read_csv(test_lst_path,sep="/",header=None, usecols=cols_to_use) ## reads lst, use / as sep to goet last column with filenames

    img_lst=img_lst.values.T.tolist()

    df = pd.DataFrame(predictions,columns = header, index=img_lst)
    df.index.name = 'image'
    
    print("Saving csv to %s" % submission_path)
    df.to_csv(submission_path)
     
    print("Compress with gzip")
    os.system("gzip -f %s" % submission_path)
    
    print("  stored in %s.gz" % submission_path)