Here I have tried to understand the data and try out multiple augmentation techniques for this competition. I will continue adding newer things to this notebook as time goes on. I hope you enjoy!
\n HAPPY WHALING !!","metadata":{}},{"cell_type":"markdown","source":"There are no weird overlaps. It's just a small spelling error. Phew
\nSo, now we can just merge these species and remove this error
","metadata":{"execution":{"iopub.status.busy":"2022-02-03T07:53:23.110406Z","iopub.execute_input":"2022-02-03T07:53:23.110702Z","iopub.status.idle":"2022-02-03T07:53:23.11702Z","shell.execute_reply.started":"2022-02-03T07:53:23.110673Z","shell.execute_reply":"2022-02-03T07:53:23.11635Z"}}},{"cell_type":"code","source":"for i in tqdm(range(train_df.shape[0])):\n if train_df.species[i] == 'kiler_whale':\n train_df.species[i] = 'killer_whale'\n elif train_df.species[i] == 'bottlenose_dolpin':\n train_df.species[i] = 'bottlenose_dolphin'","metadata":{"execution":{"iopub.status.busy":"2022-02-04T10:47:31.039817Z","iopub.execute_input":"2022-02-04T10:47:31.040500Z","iopub.status.idle":"2022-02-04T10:47:32.349874Z","shell.execute_reply.started":"2022-02-04T10:47:31.040463Z","shell.execute_reply":"2022-02-04T10:47:32.349062Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"So, now we should have only 28 unique species","metadata":{}},{"cell_type":"code","source":"train_df['species'].nunique()","metadata":{"execution":{"iopub.status.busy":"2022-02-04T10:47:32.351964Z","iopub.execute_input":"2022-02-04T10:47:32.352207Z","iopub.status.idle":"2022-02-04T10:47:32.361941Z","shell.execute_reply.started":"2022-02-04T10:47:32.352179Z","shell.execute_reply":"2022-02-04T10:47:32.361248Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"P.S.\nAfter writing this part of the code, I found this discussion here that discusses on this
","metadata":{}},{"cell_type":"markdown","source":"