{"cells":[{"metadata":{"trusted":true},"cell_type":"code","source":"library(keras)\nlibrary(tensorflow)\n#tfe_enable_eager_execution(device_policy = \"silent\")\n\nlibrary(tfdatasets)\nlibrary(stringr)\nlibrary(dplyr)\n\nuse_session_with_seed(7777,\n disable_gpu = FALSE,\n disable_parallel_cpu = FALSE)\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"audio <- tf$contrib$framework$python$ops$audio_ops\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"files <- fs::dir_ls(path = \"../input/train_curated/\",\n recursive = TRUE,\n glob = \"*.wav\")\n\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"cast_variables <- function(df, variable){\n x <- as.character(unique(df[[variable]]))\n x <- gsub(\" \", \"\", toString(x)) #so it can split on strings like \"A1,A2\" and \"A1, A2\"\n x <- unlist(strsplit(x, \",\"))\n x <- as.character(x)\n new_columns <- unique(sort(x))\n for (i in seq_along(new_columns)){\n df$temp <- NA\n df$temp <- ifelse(grepl(new_columns[i], df[[variable]]), 1, 0)\n colnames(df)[colnames(df) == \"temp\"] <- new_columns[i]\n }\n return(df)\n}\n\ntrain <- read.csv('../input/train_curated.csv', stringsAsFactors = F)\nnew_df <- cast_variables(train, 'labels')\nnew_df$fname <- paste('../input/train_curated/',train$fname, sep = \"\")\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"head(new_df)\nlabels <- c('Accelerating_and_revving_and_vroom','Accordion','Acoustic_guitar', 'Applause', 'Bark' ,'Bass_drum',\n 'Bass_guitar' ,'Bathtub_(filling_or_washing)', 'Bicycle_bell' ,'Burping_and_eructation', 'Bus' ,'Buzz' ,\n 'Car_passing_by' ,'Cheering' ,'Chewing_and_mastication', 'Child_speech_and_kid_speaking' ,\n 'Chink_and_clink', 'Chirp_and_tweet', 'Church_bell', 'Clapping', 'Computer_keyboard', 'Crackle' ,'Cricket',\n 'Crowd', 'Cupboard_open_or_close', 'Cutlery_and_silverware' ,'Dishes_and_pots_and_pans',\n 'Drawer_open_or_close', 'Drip' ,'Electric_guitar', 'Fart' ,'Female_singing',\n 'Female_speech_and_woman_speaking', 'Fill_(with_liquid)', 'Finger_snapping', 'Frying_(food)' ,'Gasp' ,\n 'Glockenspiel', 'Gong', 'Gurgling', 'Harmonica', 'Hi-hat', 'Hiss' ,'Keys_jangling', 'Knock',\n 'Male_singing', 'Male_speech_and_man_speaking', 'Marimba_and_xylophone', 'Mechanical_fan', 'Meow',\n 'Microwave_oven', 'Motorcycle' ,'Printer' ,'Purr', 'Race_car_and_auto_racing' ,'Raindrop', 'Run' ,\n 'Scissors', 'Screaming', 'Shatter', 'Sigh', 'Sink_(filling_or_washing)', 'Skateboard' ,'Slam' ,'Sneeze',\n 'Squeak', 'Stream', 'Strum', 'Tap' ,'Tick-tock', 'Toilet_flush', 'Traffic_noise_and_roadway_noise',\n 'Trickle_and_dribble', 'Walk_and_footsteps', 'Water_tap_and_faucet' ,'Waves_and_surf' ,'Whispering',\n 'Writing' ,'Yell' ,'Zipper_(clothing)')","execution_count":49,"outputs":[{"output_type":"display_data","data":{"text/plain":" fname labels \n1 ../input/train_curated/0006ae4e.wav Bark \n2 ../input/train_curated/0019ef41.wav Raindrop \n3 ../input/train_curated/001ec0ad.wav Finger_snapping\n4 ../input/train_curated/0026c7cb.wav Run \n5 ../input/train_curated/0026f116.wav Finger_snapping\n6 ../input/train_curated/003be5db.wav Whispering \n Accelerating_and_revving_and_vroom Accordion Acoustic_guitar Applause Bark\n1 0 0 0 0 1 \n2 0 0 0 0 0 \n3 0 0 0 0 0 \n4 0 0 0 0 0 \n5 0 0 0 0 0 \n6 0 0 0 0 0 \n Bass_drum Bass_guitar Bathtub_(filling_or_washing) ⋯ Toilet_flush\n1 0 0 0 ⋯ 0 \n2 0 0 0 ⋯ 0 \n3 0 0 0 ⋯ 0 \n4 0 0 0 ⋯ 0 \n5 0 0 0 ⋯ 0 \n6 0 0 0 ⋯ 0 \n Traffic_noise_and_roadway_noise Trickle_and_dribble Walk_and_footsteps\n1 0 0 0 \n2 0 0 0 \n3 0 0 0 \n4 0 0 0 \n5 0 0 0 \n6 0 0 0 \n Water_tap_and_faucet Waves_and_surf Whispering Writing Yell Zipper_(clothing)\n1 0 0 0 0 0 0 \n2 0 0 0 0 0 0 \n3 0 0 0 0 0 0 \n4 0 0 0 0 0 0 \n5 0 0 0 0 0 0 \n6 0 0 1 0 0 0 ","text/latex":"\\begin{tabular}{r|llllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllll}\n fname & labels & Accelerating\\_and\\_revving\\_and\\_vroom & Accordion & Acoustic\\_guitar & Applause & Bark & Bass\\_drum & Bass\\_guitar & Bathtub\\_(filling\\_or\\_washing) & Bicycle\\_bell & Burping\\_and\\_eructation & Bus & Buzz & Car\\_passing\\_by & Cheering & Chewing\\_and\\_mastication & Child\\_speech\\_and\\_kid\\_speaking & Chink\\_and\\_clink & Chirp\\_and\\_tweet & Church\\_bell & Clapping & Computer\\_keyboard & Crackle & Cricket & Crowd & Cupboard\\_open\\_or\\_close & Cutlery\\_and\\_silverware & Dishes\\_and\\_pots\\_and\\_pans & Drawer\\_open\\_or\\_close & Drip & Electric\\_guitar & Fart & Female\\_singing & Female\\_speech\\_and\\_woman\\_speaking & Fill\\_(with\\_liquid) & Finger\\_snapping & Frying\\_(food) & Gasp & Glockenspiel & Gong & Gurgling & Harmonica & Hi-hat & Hiss & Keys\\_jangling & Knock & Male\\_singing & Male\\_speech\\_and\\_man\\_speaking & Marimba\\_and\\_xylophone & Mechanical\\_fan & Meow & Microwave\\_oven & Motorcycle & Printer & Purr & Race\\_car\\_and\\_auto\\_racing & Raindrop & Run & Scissors & Screaming & Shatter & Sigh & Sink\\_(filling\\_or\\_washing) & Skateboard & Slam & Sneeze & Squeak & Stream & Strum & Tap & Tick-tock & Toilet\\_flush & Traffic\\_noise\\_and\\_roadway\\_noise & Trickle\\_and\\_dribble & Walk\\_and\\_footsteps & Water\\_tap\\_and\\_faucet & Waves\\_and\\_surf & Whispering & Writing & Yell & Zipper\\_(clothing)\\\\\n\\hline\n\t ../input/train\\_curated/0006ae4e.wav & Bark & 0 & 0 & 0 & 0 & 1 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\\\\n\t ../input/train\\_curated/0019ef41.wav & Raindrop & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 1 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\\\\n\t ../input/train\\_curated/001ec0ad.wav & Finger\\_snapping & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 1 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\\\\n\t ../input/train\\_curated/0026c7cb.wav & Run & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 1 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\\\\n\t ../input/train\\_curated/0026f116.wav & Finger\\_snapping & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 1 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\\\\n\t ../input/train\\_curated/003be5db.wav & Whispering & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 1 & 0 & 0 & 0 \\\\\n\\end{tabular}\n","text/markdown":"\n| fname | labels | Accelerating_and_revving_and_vroom | Accordion | Acoustic_guitar | Applause | Bark | Bass_drum | Bass_guitar | Bathtub_(filling_or_washing) | ⋯ | Toilet_flush | Traffic_noise_and_roadway_noise | Trickle_and_dribble | Walk_and_footsteps | Water_tap_and_faucet | Waves_and_surf | Whispering | Writing | Yell | Zipper_(clothing) |\n|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| ../input/train_curated/0006ae4e.wav | Bark | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | ⋯ | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |\n| ../input/train_curated/0019ef41.wav | Raindrop | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | ⋯ | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |\n| ../input/train_curated/001ec0ad.wav | Finger_snapping | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | ⋯ | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |\n| ../input/train_curated/0026c7cb.wav | Run | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | ⋯ | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |\n| ../input/train_curated/0026f116.wav | Finger_snapping | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | ⋯ | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |\n| ../input/train_curated/003be5db.wav | Whispering | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | ⋯ | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 |\n\n","text/html":"\n\n\t\n\n\n\t\n\t\n\t\n\t\n\t\n\t\n\n
fnamelabelsAccelerating_and_revving_and_vroomAccordionAcoustic_guitarApplauseBarkBass_drumBass_guitarBathtub_(filling_or_washing)⋯Toilet_flushTraffic_noise_and_roadway_noiseTrickle_and_dribbleWalk_and_footstepsWater_tap_and_faucetWaves_and_surfWhisperingWritingYellZipper_(clothing)
../input/train_curated/0006ae4e.wavBark 0 0 0 0 1 0 0 0 ⋯ 0 0 0 0 0 0 0 0 0 0
../input/train_curated/0019ef41.wavRaindrop 0 0 0 0 0 0 0 0 ⋯ 0 0 0 0 0 0 0 0 0 0
../input/train_curated/001ec0ad.wavFinger_snapping 0 0 0 0 0 0 0 0 ⋯ 0 0 0 0 0 0 0 0 0 0
../input/train_curated/0026c7cb.wavRun 0 0 0 0 0 0 0 0 ⋯ 0 0 0 0 0 0 0 0 0 0
../input/train_curated/0026f116.wavFinger_snapping 0 0 0 0 0 0 0 0 ⋯ 0 0 0 0 0 0 0 0 0 0
../input/train_curated/003be5db.wavWhispering 0 0 0 0 0 0 0 0 ⋯ 0 0 0 0 0 0 1 0 0 0
\n"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"new_df[1,labels]","execution_count":50,"outputs":[{"output_type":"display_data","data":{"text/plain":" Accelerating_and_revving_and_vroom Accordion Acoustic_guitar Applause Bark\n1 0 0 0 0 1 \n Bass_drum Bass_guitar Bathtub_(filling_or_washing) Bicycle_bell\n1 0 0 0 0 \n Burping_and_eructation ⋯ Toilet_flush Traffic_noise_and_roadway_noise\n1 0 ⋯ 0 0 \n Trickle_and_dribble Walk_and_footsteps Water_tap_and_faucet Waves_and_surf\n1 0 0 0 0 \n Whispering Writing Yell Zipper_(clothing)\n1 0 0 0 0 ","text/latex":"\\begin{tabular}{r|llllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllll}\n Accelerating\\_and\\_revving\\_and\\_vroom & Accordion & Acoustic\\_guitar & Applause & Bark & Bass\\_drum & Bass\\_guitar & Bathtub\\_(filling\\_or\\_washing) & Bicycle\\_bell & Burping\\_and\\_eructation & Bus & Buzz & Car\\_passing\\_by & Cheering & Chewing\\_and\\_mastication & Child\\_speech\\_and\\_kid\\_speaking & Chink\\_and\\_clink & Chirp\\_and\\_tweet & Church\\_bell & Clapping & Computer\\_keyboard & Crackle & Cricket & Crowd & Cupboard\\_open\\_or\\_close & Cutlery\\_and\\_silverware & Dishes\\_and\\_pots\\_and\\_pans & Drawer\\_open\\_or\\_close & Drip & Electric\\_guitar & Fart & Female\\_singing & Female\\_speech\\_and\\_woman\\_speaking & Fill\\_(with\\_liquid) & Finger\\_snapping & Frying\\_(food) & Gasp & Glockenspiel & Gong & Gurgling & Harmonica & Hi-hat & Hiss & Keys\\_jangling & Knock & Male\\_singing & Male\\_speech\\_and\\_man\\_speaking & Marimba\\_and\\_xylophone & Mechanical\\_fan & Meow & Microwave\\_oven & Motorcycle & Printer & Purr & Race\\_car\\_and\\_auto\\_racing & Raindrop & Run & Scissors & Screaming & Shatter & Sigh & Sink\\_(filling\\_or\\_washing) & Skateboard & Slam & Sneeze & Squeak & Stream & Strum & Tap & Tick-tock & Toilet\\_flush & Traffic\\_noise\\_and\\_roadway\\_noise & Trickle\\_and\\_dribble & Walk\\_and\\_footsteps & Water\\_tap\\_and\\_faucet & Waves\\_and\\_surf & Whispering & Writing & Yell & Zipper\\_(clothing)\\\\\n\\hline\n\t 0 & 0 & 0 & 0 & 1 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0\\\\\n\\end{tabular}\n","text/markdown":"\n| Accelerating_and_revving_and_vroom | Accordion | Acoustic_guitar | Applause | Bark | Bass_drum | Bass_guitar | Bathtub_(filling_or_washing) | Bicycle_bell | Burping_and_eructation | ⋯ | Toilet_flush | Traffic_noise_and_roadway_noise | Trickle_and_dribble | Walk_and_footsteps | Water_tap_and_faucet | Waves_and_surf | Whispering | Writing | Yell | Zipper_(clothing) |\n|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | ⋯ | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |\n\n","text/html":"\n\n\t\n\n\n\t\n\n
Accelerating_and_revving_and_vroomAccordionAcoustic_guitarApplauseBarkBass_drumBass_guitarBathtub_(filling_or_washing)Bicycle_bellBurping_and_eructation⋯Toilet_flushTraffic_noise_and_roadway_noiseTrickle_and_dribbleWalk_and_footstepsWater_tap_and_faucetWaves_and_surfWhisperingWritingYellZipper_(clothing)
0000100000⋯0000000000
\n"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"batch_size <- 128\nbuffer_size <- nrow(new_df)\n\nwindow_size_ms <- 30L \nwindow_stride_ms <- 10L \n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"data_generator <- function(df, window_size_ms, window_stride_ms, shuffle = FALSE) {\n \n sampling_rate <- 44100\n window_size <- as.integer(sampling_rate*window_size_ms/1000)\n stride <- as.integer(sampling_rate*window_stride_ms/1000)\n #fft_size <- as.integer(2^trunc(log(window_size, 2)) + 1)\n #n_chunks <- length(seq(window_size/2, sampling_rate - window_size/2, stride))\n \n ds <- tensor_slices_dataset(df)\n \n if(shuffle)\n ds <- ds %>% dataset_shuffle(buffer_size = 100) \n \n ds <- ds %>%\n dataset_map(function(obs) {\n \n # decoding wav files\n audio_binary <- tf$read_file(tf$reshape(obs$fname, shape = list()))\n wav <- audio$decode_wav(audio_binary, desired_channels = 1)\n \n # create the spectrogram\n spectrogram <- audio$audio_spectrogram(\n wav$audio, \n window_size = window_size, \n stride = stride,\n magnitude_squared = TRUE\n )\n \n # Custom brightness\n #brightness = tf$placeholder(tf$float32, shape=c())\n brightness <- 100L\n mul = tf$multiply(spectrogram, brightness)\n \n # Normalize pixels\n min_const = tf$constant(255)\n minimum = tf$minimum(mul, min_const)\n \n # Expand dims so we get the proper shape\n expand_dims = tf$expand_dims(minimum, -1L)\n \n # Resize the spectrogram to input size of the model\n resize = tf$image$resize_bilinear(expand_dims, c(128L, 128L))\n \n # Remove the trailing dimension\n squeeze = tf$squeeze(resize, 0L)\n \n # Tensorflow spectrogram has time along y axis and frequencies along x axis\n # so we fix that\n #flip = tf$image$flip_left_right(squeeze)\n #transpose = tf$image$transpose_image(flip)\n \n # Convert image to 3 channels, it's still a grayscale image however\n #grayscale = tf$image$grayscale_to_rgb(transpose)\n \n # Cast to uint8 and encode as png\n #cast = tf$cast(grayscale, tf$uint8)\n #spectrogram <- tf$log(tf$abs(spectrogram) + 0.01)\n #spectrogram <- tf$transpose(spectrogram, perm = c(1L, 2L, 0L))\n \n # transform the class_id into a one-hot encoded vector\n response <- as.matrix(obs[labels])\n list(squeeze, response)\n }) %>%\n dataset_repeat()\n \n ds <- ds %>% \n dataset_padded_batch(batch_size, list(shape(128, 128, 1), shape(80)))\n \n return(ds) \n}\n\n","execution_count":68,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"id_train <- sample(nrow(new_df), size = 0.7 * nrow(new_df))\nn_train <- length(id_train)\nn_test <- nrow(new_df) - n_train\n\nds_train <- data_generator(new_df[id_train, ],\n window_size_ms = window_size_ms,\n window_stride_ms = window_stride_ms)\nds_val <- data_generator(new_df[-id_train, ],\n window_size_ms = window_size_ms,\n window_stride_ms = window_stride_ms)\n","execution_count":69,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model <- keras_model_sequential()\nmodel %>% \n layer_conv_2d(input_shape = c(128, 128, 1), \n filters = 32, kernel_size = c(3,3), activation = 'relu') %>% \n layer_max_pooling_2d(pool_size = c(2, 2)) %>% \n layer_conv_2d(filters = 64, kernel_size = c(3,3), activation = 'relu') %>% \n layer_max_pooling_2d(pool_size = c(2, 2)) %>% \n layer_conv_2d(filters = 128, kernel_size = c(3,3), activation = 'relu') %>% \n layer_max_pooling_2d(pool_size = c(2, 2)) %>% \n layer_conv_2d(filters = 256, kernel_size = c(3,3), activation = 'relu') %>% \n layer_max_pooling_2d(pool_size = c(2, 2)) %>% \n layer_dropout(rate = 0.25) %>% \n layer_flatten() %>% \n layer_dense(units = 128, activation = 'relu') %>% \n layer_dropout(rate = 0.2) %>% \n layer_dense(units = 80, activation = 'softmax')\n\nmodel %>% compile(\n loss = loss_binary_crossentropy,\n optimizer = optimizer_adam(),\n metrics = c('accuracy')\n)\nmodel %>% summary()\n","execution_count":53,"outputs":[{"output_type":"stream","text":"________________________________________________________________________________\nLayer (type) Output Shape Param # \n================================================================================\nconv2d_1 (Conv2D) (None, 126, 126, 32) 320 \n________________________________________________________________________________\nmax_pooling2d_1 (MaxPooling2D) (None, 63, 63, 32) 0 \n________________________________________________________________________________\nconv2d_2 (Conv2D) (None, 61, 61, 64) 18496 \n________________________________________________________________________________\nmax_pooling2d_2 (MaxPooling2D) (None, 30, 30, 64) 0 \n________________________________________________________________________________\nconv2d_3 (Conv2D) (None, 28, 28, 128) 73856 \n________________________________________________________________________________\nmax_pooling2d_3 (MaxPooling2D) (None, 14, 14, 128) 0 \n________________________________________________________________________________\nconv2d_4 (Conv2D) (None, 12, 12, 256) 295168 \n________________________________________________________________________________\nmax_pooling2d_4 (MaxPooling2D) (None, 6, 6, 256) 0 \n________________________________________________________________________________\ndropout_1 (Dropout) (None, 6, 6, 256) 0 \n________________________________________________________________________________\nflatten_1 (Flatten) (None, 9216) 0 \n________________________________________________________________________________\ndense_1 (Dense) (None, 128) 1179776 \n________________________________________________________________________________\ndropout_2 (Dropout) (None, 128) 0 \n________________________________________________________________________________\ndense_2 (Dense) (None, 80) 10320 \n================================================================================\nTotal params: 1,577,936\nTrainable params: 1,577,936\nNon-trainable params: 0\n________________________________________________________________________________\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"model %>% fit_generator(\n generator = ds_train,\n steps_per_epoch = 0.7*nrow(new_df)/batch_size,\n epochs = 25,\n verbose = 1,\n validation_data = ds_val, \n validation_steps = 0.3*nrow(new_df)/batch_size\n)\n","execution_count":70,"outputs":[{"output_type":"stream","text":"Warning message in normalizePath(path):\n“path[1]=\"\": No such file or directory”Warning message in normalizePath(path):\n“path[1]=\"\": No such file or directory”","name":"stderr"},{"output_type":"error","ename":"ERROR","evalue":"Error in py_call_impl(callable, dots$args, dots$keywords): TypeError: 'NoneType' object is not callable\n\nDetailed traceback: \n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/keras/legacy/interfaces.py\", line 91, in wrapper\n return func(*args, **kwargs)\n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/keras/engine/training.py\", line 1418, in fit_generator\n initial_epoch=initial_epoch)\n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/keras/engine/training_generator.py\", line 181, in fit_generator\n generator_output = next(output_generator)\n File \"/usr/local/lib/R/site-library/keras/python/kerastools/generator.py\", line 11, in gen\n yield session.run(batch)\n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 877, in run\n run_metadata_ptr)\n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1100, in _run\n feed_dict_tensor, options, run_metadata)\n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1272, in _do_run\n run_metadata)\n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1278, in _do_call\n return fn(*args)\n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1263, in _run_fn\n options, feed_dict, fetch_list, target_list, run_metadata)\n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1350, in _call_tf_sessionrun\n run_metadata)\n\n","traceback":["Error in py_call_impl(callable, dots$args, dots$keywords): TypeError: 'NoneType' object is not callable\n\nDetailed traceback: \n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/keras/legacy/interfaces.py\", line 91, in wrapper\n return func(*args, **kwargs)\n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/keras/engine/training.py\", line 1418, in fit_generator\n initial_epoch=initial_epoch)\n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/keras/engine/training_generator.py\", line 181, in fit_generator\n generator_output = next(output_generator)\n File \"/usr/local/lib/R/site-library/keras/python/kerastools/generator.py\", line 11, in gen\n yield session.run(batch)\n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 877, in run\n run_metadata_ptr)\n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1100, in _run\n feed_dict_tensor, options, run_metadata)\n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1272, in _do_run\n run_metadata)\n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1278, in _do_call\n return fn(*args)\n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1263, in _run_fn\n options, feed_dict, fetch_list, target_list, run_metadata)\n File \"/root/.virtualenvs/r-tensorflow/lib/python2.7/site-packages/tensorflow/python/client/session.py\", line 1350, in _call_tf_sessionrun\n run_metadata)\n\nTraceback:\n","1. model %>% fit_generator(generator = ds_train, steps_per_epoch = 0.7 * \n . nrow(new_df)/32, epochs = 1, verbose = 1, validation_data = ds_val, \n . validation_steps = 0.3 * nrow(new_df)/32)","2. withVisible(eval(quote(`_fseq`(`_lhs`)), env, env))","3. eval(quote(`_fseq`(`_lhs`)), env, env)","4. eval(quote(`_fseq`(`_lhs`)), env, env)","5. `_fseq`(`_lhs`)","6. freduce(value, `_function_list`)","7. withVisible(function_list[[k]](value))","8. function_list[[k]](value)","9. fit_generator(., generator = ds_train, steps_per_epoch = 0.7 * \n . nrow(new_df)/32, epochs = 1, verbose = 1, validation_data = ds_val, \n . validation_steps = 0.3 * nrow(new_df)/32)","10. call_generator_function(object$fit_generator, list(generator = generator, \n . steps_per_epoch = as.integer(steps_per_epoch), epochs = as.integer(epochs), \n . verbose = as.integer(verbose), callbacks = normalize_callbacks(view_metrics, \n . callbacks), validation_data = validation_data, validation_steps = as_nullable_integer(validation_steps), \n . class_weight = as_class_weight(class_weight), max_queue_size = as.integer(max_queue_size), \n . workers = as.integer(workers), initial_epoch = as.integer(initial_epoch)))","11. do.call(func, args)","12. (structure(function (...) \n . {\n . dots <- py_resolve_dots(list(...))\n . result <- py_call_impl(callable, dots$args, dots$keywords)\n . if (convert) {\n . result <- py_to_r(result)\n . if (is.null(result)) \n . invisible(result)\n . else result\n . }\n . else {\n . result\n . }\n . }, class = c(\"python.builtin.instancemethod\", \"python.builtin.object\"\n . ), py_object = ))(generator = , steps_per_epoch = 108L, \n . epochs = 1L, verbose = 1L, callbacks = list(), \n . validation_data = , validation_steps = 46L, \n . class_weight = NULL, max_queue_size = 10L, workers = 0L, \n . initial_epoch = 0L, use_multiprocessing = FALSE)","13. py_call_impl(callable, dots$args, dots$keywords)"]}]},{"metadata":{"trusted":true},"cell_type":"code","source":"test <- read.csv('../input/sample_submission.csv', stringsAsFactors = F)\ncnames <- data.frame(fname = test$fname)\ntest$fname <- paste('../input/test/',test$fname, sep = \"\")\ncolnames(test) <- c('fname', labels)\n\n","execution_count":91,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"cnames","execution_count":93,"outputs":[{"output_type":"display_data","data":{"text/plain":" fname \n1 000ccb97.wav\n2 0012633b.wav\n3 001ed5f1.wav\n4 00294be0.wav\n5 003fde7a.wav\n6 0040ccc9.wav\n7 0046b732.wav\n8 004f3bbc.wav\n9 00526050.wav\n10 00559da4.wav\n11 00582bbe.wav\n12 0064aedf.wav\n13 0065512b.wav\n14 006a91d2.wav\n15 006ea9ee.wav\n16 006f9dca.wav\n17 007450dc.wav\n18 00979c8a.wav\n19 00992464.wav\n20 00b44a8a.wav\n21 00bfaaaf.wav\n22 00cf9680.wav\n23 00eae94d.wav\n24 011ec5b8.wav\n25 013200dc.wav\n26 01440c2e.wav\n27 01567a3d.wav\n28 0163d203.wav\n29 017c24b2.wav\n30 01a992c5.wav\n⋮ ⋮ \n1091 406306b6.wav\n1092 406a1cc1.wav\n1093 406ade5e.wav\n1094 4093913f.wav\n1095 40957335.wav\n1096 4098028d.wav\n1097 40afd1b3.wav\n1098 40c1ed87.wav\n1099 40c50046.wav\n1100 40d0726d.wav\n1101 40dd1274.wav\n1102 40e25750.wav\n1103 410837eb.wav\n1104 410c2138.wav\n1105 411993fd.wav\n1106 413b5427.wav\n1107 41522a65.wav\n1108 415f4c8f.wav\n1109 41759973.wav\n1110 419271cc.wav\n1111 4195cb1a.wav\n1112 41af688a.wav\n1113 41c60bba.wav\n1114 41cfee6d.wav\n1115 41f16193.wav\n1116 41f1ea4a.wav\n1117 41f86bc4.wav\n1118 4215309a.wav\n1119 4248d196.wav\n1120 42542036.wav","text/latex":"\\begin{tabular}{r|l}\n fname\\\\\n\\hline\n\t 000ccb97.wav\\\\\n\t 0012633b.wav\\\\\n\t 001ed5f1.wav\\\\\n\t 00294be0.wav\\\\\n\t 003fde7a.wav\\\\\n\t 0040ccc9.wav\\\\\n\t 0046b732.wav\\\\\n\t 004f3bbc.wav\\\\\n\t 00526050.wav\\\\\n\t 00559da4.wav\\\\\n\t 00582bbe.wav\\\\\n\t 0064aedf.wav\\\\\n\t 0065512b.wav\\\\\n\t 006a91d2.wav\\\\\n\t 006ea9ee.wav\\\\\n\t 006f9dca.wav\\\\\n\t 007450dc.wav\\\\\n\t 00979c8a.wav\\\\\n\t 00992464.wav\\\\\n\t 00b44a8a.wav\\\\\n\t 00bfaaaf.wav\\\\\n\t 00cf9680.wav\\\\\n\t 00eae94d.wav\\\\\n\t 011ec5b8.wav\\\\\n\t 013200dc.wav\\\\\n\t 01440c2e.wav\\\\\n\t 01567a3d.wav\\\\\n\t 0163d203.wav\\\\\n\t 017c24b2.wav\\\\\n\t 01a992c5.wav\\\\\n\t ⋮\\\\\n\t 406306b6.wav\\\\\n\t 406a1cc1.wav\\\\\n\t 406ade5e.wav\\\\\n\t 4093913f.wav\\\\\n\t 40957335.wav\\\\\n\t 4098028d.wav\\\\\n\t 40afd1b3.wav\\\\\n\t 40c1ed87.wav\\\\\n\t 40c50046.wav\\\\\n\t 40d0726d.wav\\\\\n\t 40dd1274.wav\\\\\n\t 40e25750.wav\\\\\n\t 410837eb.wav\\\\\n\t 410c2138.wav\\\\\n\t 411993fd.wav\\\\\n\t 413b5427.wav\\\\\n\t 41522a65.wav\\\\\n\t 415f4c8f.wav\\\\\n\t 41759973.wav\\\\\n\t 419271cc.wav\\\\\n\t 4195cb1a.wav\\\\\n\t 41af688a.wav\\\\\n\t 41c60bba.wav\\\\\n\t 41cfee6d.wav\\\\\n\t 41f16193.wav\\\\\n\t 41f1ea4a.wav\\\\\n\t 41f86bc4.wav\\\\\n\t 4215309a.wav\\\\\n\t 4248d196.wav\\\\\n\t 42542036.wav\\\\\n\\end{tabular}\n","text/markdown":"\n| fname |\n|---|\n| 000ccb97.wav |\n| 0012633b.wav |\n| 001ed5f1.wav |\n| 00294be0.wav |\n| 003fde7a.wav |\n| 0040ccc9.wav |\n| 0046b732.wav |\n| 004f3bbc.wav |\n| 00526050.wav |\n| 00559da4.wav |\n| 00582bbe.wav |\n| 0064aedf.wav |\n| 0065512b.wav |\n| 006a91d2.wav |\n| 006ea9ee.wav |\n| 006f9dca.wav |\n| 007450dc.wav |\n| 00979c8a.wav |\n| 00992464.wav |\n| 00b44a8a.wav |\n| 00bfaaaf.wav |\n| 00cf9680.wav |\n| 00eae94d.wav |\n| 011ec5b8.wav |\n| 013200dc.wav |\n| 01440c2e.wav |\n| 01567a3d.wav |\n| 0163d203.wav |\n| 017c24b2.wav |\n| 01a992c5.wav |\n| ⋮ |\n| 406306b6.wav |\n| 406a1cc1.wav |\n| 406ade5e.wav |\n| 4093913f.wav |\n| 40957335.wav |\n| 4098028d.wav |\n| 40afd1b3.wav |\n| 40c1ed87.wav |\n| 40c50046.wav |\n| 40d0726d.wav |\n| 40dd1274.wav |\n| 40e25750.wav |\n| 410837eb.wav |\n| 410c2138.wav |\n| 411993fd.wav |\n| 413b5427.wav |\n| 41522a65.wav |\n| 415f4c8f.wav |\n| 41759973.wav |\n| 419271cc.wav |\n| 4195cb1a.wav |\n| 41af688a.wav |\n| 41c60bba.wav |\n| 41cfee6d.wav |\n| 41f16193.wav |\n| 41f1ea4a.wav |\n| 41f86bc4.wav |\n| 4215309a.wav |\n| 4248d196.wav |\n| 42542036.wav |\n\n","text/html":"\n\n\t\n\n\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\n
fname
000ccb97.wav
0012633b.wav
001ed5f1.wav
00294be0.wav
003fde7a.wav
0040ccc9.wav
0046b732.wav
004f3bbc.wav
00526050.wav
00559da4.wav
00582bbe.wav
0064aedf.wav
0065512b.wav
006a91d2.wav
006ea9ee.wav
006f9dca.wav
007450dc.wav
00979c8a.wav
00992464.wav
00b44a8a.wav
00bfaaaf.wav
00cf9680.wav
00eae94d.wav
011ec5b8.wav
013200dc.wav
01440c2e.wav
01567a3d.wav
0163d203.wav
017c24b2.wav
01a992c5.wav
⋮
406306b6.wav
406a1cc1.wav
406ade5e.wav
4093913f.wav
40957335.wav
4098028d.wav
40afd1b3.wav
40c1ed87.wav
40c50046.wav
40d0726d.wav
40dd1274.wav
40e25750.wav
410837eb.wav
410c2138.wav
411993fd.wav
413b5427.wav
41522a65.wav
415f4c8f.wav
41759973.wav
419271cc.wav
4195cb1a.wav
41af688a.wav
41c60bba.wav
41cfee6d.wav
41f16193.wav
41f1ea4a.wav
41f86bc4.wav
4215309a.wav
4248d196.wav
42542036.wav
\n"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"ds_test <- data_generator(test,\n window_size_ms = window_size_ms,\n window_stride_ms = window_stride_ms)\n","execution_count":77,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"n_steps <- nrow(test)/batch_size +1\npredictions <- predict_generator(object = model,\n ds_test, \n steps = n_steps\n)","execution_count":97,"outputs":[{"output_type":"stream","text":"Warning message in normalizePath(path):\n“path[1]=\"\": No such file or directory”","name":"stderr"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"predictions <- as.data.frame(predictions[1:1120,])\n\npredictions <- cbind(cnames, predictions)\n","execution_count":98,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"colnames(predictions) <- c('fname', labels)\nhead(predictions)","execution_count":99,"outputs":[{"output_type":"display_data","data":{"text/plain":" fname Accelerating_and_revving_and_vroom Accordion Acoustic_guitar\n1 000ccb97.wav 0.013209159 0.01774904 0.002916300 \n2 0012633b.wav 0.017934894 0.01739276 0.008245957 \n3 001ed5f1.wav 0.014889697 0.01000734 0.014423068 \n4 00294be0.wav 0.023437565 0.01410034 0.012042203 \n5 003fde7a.wav 0.005750209 0.01126031 0.003639374 \n6 0040ccc9.wav 0.014303586 0.01032966 0.024132181 \n Applause Bark Bass_drum Bass_guitar Bathtub_(filling_or_washing)\n1 0.009029981 0.019333581 0.009261215 0.01130432 0.003191851 \n2 0.005662751 0.019713445 0.018844701 0.02341015 0.005488900 \n3 0.012587036 0.008792087 0.010042071 0.01154019 0.010289351 \n4 0.005482092 0.023469465 0.023048503 0.03446521 0.004114904 \n5 0.006456756 0.008438641 0.006241641 0.01137754 0.003585051 \n6 0.007142874 0.026701186 0.026633421 0.04539999 0.003292581 \n Bicycle_bell ⋯ Toilet_flush Traffic_noise_and_roadway_noise\n1 0.041688547 ⋯ 0.021906769 0.004111980 \n2 0.009091829 ⋯ 0.007201934 0.013194967 \n3 0.017354663 ⋯ 0.015117173 0.010329596 \n4 0.007335012 ⋯ 0.004856232 0.016843693 \n5 0.021493951 ⋯ 0.016109822 0.004421263 \n6 0.006904302 ⋯ 0.004810598 0.014251505 \n Trickle_and_dribble Walk_and_footsteps Water_tap_and_faucet Waves_and_surf\n1 0.011314391 0.01709291 0.005540514 0.001942767 \n2 0.005653424 0.01434080 0.012078281 0.013583345 \n3 0.011478883 0.01193406 0.008192242 0.012718624 \n4 0.004171766 0.01782716 0.008663573 0.019125421 \n5 0.009133413 0.01445377 0.006500922 0.005644536 \n6 0.002832420 0.02637555 0.005835421 0.012582770 \n Whispering Writing Yell Zipper_(clothing)\n1 0.013837142 0.003929277 0.002548119 0.002510201 \n2 0.006861777 0.010532438 0.009440523 0.007197834 \n3 0.009315304 0.012257267 0.015921116 0.016205560 \n4 0.007035829 0.014549707 0.007871445 0.009078503 \n5 0.026017219 0.013965773 0.034930632 0.007944757 \n6 0.007445882 0.016933151 0.005157431 0.006874123 ","text/latex":"\\begin{tabular}{r|lllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllll}\n fname & Accelerating\\_and\\_revving\\_and\\_vroom & Accordion & Acoustic\\_guitar & Applause & Bark & Bass\\_drum & Bass\\_guitar & Bathtub\\_(filling\\_or\\_washing) & Bicycle\\_bell & Burping\\_and\\_eructation & Bus & Buzz & Car\\_passing\\_by & Cheering & Chewing\\_and\\_mastication & Child\\_speech\\_and\\_kid\\_speaking & Chink\\_and\\_clink & Chirp\\_and\\_tweet & Church\\_bell & Clapping & Computer\\_keyboard & Crackle & Cricket & Crowd & Cupboard\\_open\\_or\\_close & Cutlery\\_and\\_silverware & Dishes\\_and\\_pots\\_and\\_pans & Drawer\\_open\\_or\\_close & Drip & Electric\\_guitar & Fart & Female\\_singing & Female\\_speech\\_and\\_woman\\_speaking & Fill\\_(with\\_liquid) & Finger\\_snapping & Frying\\_(food) & Gasp & Glockenspiel & Gong & Gurgling & Harmonica & Hi-hat & Hiss & Keys\\_jangling & Knock & Male\\_singing & Male\\_speech\\_and\\_man\\_speaking & Marimba\\_and\\_xylophone & Mechanical\\_fan & Meow & Microwave\\_oven & Motorcycle & Printer & Purr & Race\\_car\\_and\\_auto\\_racing & Raindrop & Run & Scissors & Screaming & Shatter & Sigh & Sink\\_(filling\\_or\\_washing) & Skateboard & Slam & Sneeze & Squeak & Stream & Strum & Tap & Tick-tock & Toilet\\_flush & Traffic\\_noise\\_and\\_roadway\\_noise & Trickle\\_and\\_dribble & Walk\\_and\\_footsteps & Water\\_tap\\_and\\_faucet & Waves\\_and\\_surf & Whispering & Writing & Yell & Zipper\\_(clothing)\\\\\n\\hline\n\t 000ccb97.wav & 0.013209159 & 0.01774904 & 0.002916300 & 0.009029981 & 0.019333581 & 0.009261215 & 0.01130432 & 0.003191851 & 0.041688547 & 0.01616487 & 0.006208298 & 0.001192837 & 0.007800685 & 0.005163141 & 0.004235446 & 0.018286170 & 0.022979250 & 0.003123493 & 0.001811213 & 0.008763734 & 0.005455693 & 0.004637791 & 0.023563787 & 0.014602840 & 0.007122426 & 0.077431396 & 0.01600173 & 0.009482731 & 0.003752868 & 0.015517743 & 0.002610853 & 0.006451949 & 0.02192709 & 0.022008033 & 0.032564986 & 0.004087019 & 0.015034988 & 0.011910632 & 0.011105557 & 0.002582513 & 0.027790910 & 0.070479862 & 0.009690703 & 0.033134036 & 0.004014829 & 0.005144641 & 0.005889448 & 0.005124948 & 0.002306624 & 0.003601131 & 0.003854264 & 0.003238718 & 0.01321410 & 0.010584972 & 0.007717168 & 0.007077281 & 0.003224410 & 0.016058266 & 0.004322215 & 0.024555588 & 0.036006674 & 0.002451096 & 0.007733088 & 0.012666701 & 0.015350062 & 0.020788360 & 0.008073686 & 0.003558831 & 0.015092493 & 0.001255016 & 0.021906769 & 0.004111980 & 0.011314391 & 0.01709291 & 0.005540514 & 0.001942767 & 0.013837142 & 0.003929277 & 0.002548119 & 0.002510201 \\\\\n\t 0012633b.wav & 0.017934894 & 0.01739276 & 0.008245957 & 0.005662751 & 0.019713445 & 0.018844701 & 0.02341015 & 0.005488900 & 0.009091829 & 0.01110286 & 0.010017196 & 0.023893131 & 0.028702613 & 0.008446265 & 0.008058712 & 0.014419679 & 0.006853271 & 0.011346299 & 0.018398978 & 0.008956402 & 0.013264019 & 0.011312488 & 0.007357309 & 0.009461206 & 0.008707073 & 0.007497184 & 0.01244946 & 0.014136061 & 0.008132869 & 0.016761061 & 0.017222064 & 0.013981271 & 0.01621240 & 0.008958937 & 0.006337418 & 0.013183136 & 0.010676935 & 0.007661374 & 0.027559306 & 0.010803664 & 0.007499687 & 0.007555198 & 0.020660453 & 0.007198415 & 0.016077917 & 0.016439976 & 0.004839256 & 0.013097009 & 0.016701907 & 0.016016098 & 0.019286741 & 0.013349412 & 0.01998162 & 0.022600425 & 0.008319237 & 0.015155503 & 0.011710194 & 0.008978957 & 0.006421956 & 0.015811652 & 0.006168193 & 0.011904058 & 0.008941093 & 0.011679389 & 0.007901901 & 0.006634920 & 0.016914099 & 0.021309281 & 0.012257265 & 0.010846834 & 0.007201934 & 0.013194967 & 0.005653424 & 0.01434080 & 0.012078281 & 0.013583345 & 0.006861777 & 0.010532438 & 0.009440523 & 0.007197834 \\\\\n\t 001ed5f1.wav & 0.014889697 & 0.01000734 & 0.014423068 & 0.012587036 & 0.008792087 & 0.010042071 & 0.01154019 & 0.010289351 & 0.017354663 & 0.01238269 & 0.014260820 & 0.011964301 & 0.010488351 & 0.016064158 & 0.014737913 & 0.007432333 & 0.010111255 & 0.016202275 & 0.016077366 & 0.017305441 & 0.008135349 & 0.009779415 & 0.013572582 & 0.019559197 & 0.013695079 & 0.008496306 & 0.01539851 & 0.008504216 & 0.010395436 & 0.013066684 & 0.013947759 & 0.011740626 & 0.01074616 & 0.016980305 & 0.008714594 & 0.011762088 & 0.013081432 & 0.012752635 & 0.013978678 & 0.013542175 & 0.010618812 & 0.010999147 & 0.015454285 & 0.012589674 & 0.013154701 & 0.015081092 & 0.011467987 & 0.011154671 & 0.018497987 & 0.012416708 & 0.010085844 & 0.011537371 & 0.01063192 & 0.007937283 & 0.016072102 & 0.010702000 & 0.014225417 & 0.008309362 & 0.011931220 & 0.009882022 & 0.013180921 & 0.015155540 & 0.007583323 & 0.014929327 & 0.020859601 & 0.012231631 & 0.017985804 & 0.008614746 & 0.008770337 & 0.009667718 & 0.015117173 & 0.010329596 & 0.011478883 & 0.01193406 & 0.008192242 & 0.012718624 & 0.009315304 & 0.012257267 & 0.015921116 & 0.016205560 \\\\\n\t 00294be0.wav & 0.023437565 & 0.01410034 & 0.012042203 & 0.005482092 & 0.023469465 & 0.023048503 & 0.03446521 & 0.004114904 & 0.007335012 & 0.01019442 & 0.013758400 & 0.033053476 & 0.028036490 & 0.006333001 & 0.007030853 & 0.013332581 & 0.007332225 & 0.010224531 & 0.017798053 & 0.008495482 & 0.013534676 & 0.013600533 & 0.004518735 & 0.005310874 & 0.007860234 & 0.008813389 & 0.01080730 & 0.012263839 & 0.009480231 & 0.012882064 & 0.012490769 & 0.012757791 & 0.01384262 & 0.009206623 & 0.005398977 & 0.009687904 & 0.012669750 & 0.005065848 & 0.033184182 & 0.006643392 & 0.007131244 & 0.011384564 & 0.016294347 & 0.006318660 & 0.018121189 & 0.015769981 & 0.004474476 & 0.012857059 & 0.013071152 & 0.016540313 & 0.020642856 & 0.012405249 & 0.01869415 & 0.027333410 & 0.004491779 & 0.016997661 & 0.010321607 & 0.007833610 & 0.005212857 & 0.012287161 & 0.004366430 & 0.012973196 & 0.008369882 & 0.010941757 & 0.006241411 & 0.006915945 & 0.012526098 & 0.023774538 & 0.012804964 & 0.009774634 & 0.004856232 & 0.016843693 & 0.004171766 & 0.01782716 & 0.008663573 & 0.019125421 & 0.007035829 & 0.014549707 & 0.007871445 & 0.009078503 \\\\\n\t 003fde7a.wav & 0.005750209 & 0.01126031 & 0.003639374 & 0.006456756 & 0.008438641 & 0.006241641 & 0.01137754 & 0.003585051 & 0.021493951 & 0.01732456 & 0.006786077 & 0.002377002 & 0.003871226 & 0.016522873 & 0.010240369 & 0.009367817 & 0.017591037 & 0.018697713 & 0.014192419 & 0.013476104 & 0.014690049 & 0.005436007 & 0.032800980 & 0.017815387 & 0.018655252 & 0.018515473 & 0.02372532 & 0.010674392 & 0.005516301 & 0.019427575 & 0.005249026 & 0.005398681 & 0.01836607 & 0.020926815 & 0.017294582 & 0.004409757 & 0.007502428 & 0.024738479 & 0.004918906 & 0.014579942 & 0.016986292 & 0.024831822 & 0.015341538 & 0.034124698 & 0.004938846 & 0.016062958 & 0.008111160 & 0.011159431 & 0.005970858 & 0.014818951 & 0.014313653 & 0.008054906 & 0.03171780 & 0.015228144 & 0.006424936 & 0.007771051 & 0.004871333 & 0.008037859 & 0.005975999 & 0.012854179 & 0.023895824 & 0.004746894 & 0.002909617 & 0.006521902 & 0.016431291 & 0.013096748 & 0.009053710 & 0.006520014 & 0.008246661 & 0.002526731 & 0.016109822 & 0.004421263 & 0.009133413 & 0.01445377 & 0.006500922 & 0.005644536 & 0.026017219 & 0.013965773 & 0.034930632 & 0.007944757 \\\\\n\t 0040ccc9.wav & 0.014303586 & 0.01032966 & 0.024132181 & 0.007142874 & 0.026701186 & 0.026633421 & 0.04539999 & 0.003292581 & 0.006904302 & 0.01063480 & 0.012400460 & 0.039836552 & 0.021951774 & 0.005834328 & 0.007612076 & 0.015978070 & 0.005338655 & 0.007342060 & 0.032187872 & 0.006541337 & 0.011004990 & 0.016777907 & 0.004644545 & 0.004136105 & 0.007314563 & 0.008546199 & 0.01847145 & 0.011603474 & 0.006263289 & 0.009635698 & 0.018242324 & 0.017696483 & 0.00972394 & 0.011206630 & 0.005926467 & 0.007258825 & 0.013852608 & 0.006473572 & 0.037145950 & 0.005456083 & 0.005142245 & 0.008982838 & 0.016431144 & 0.005062030 & 0.010535703 & 0.014244744 & 0.003384971 & 0.014074939 & 0.008883812 & 0.009444861 & 0.018275788 & 0.014280517 & 0.01570508 & 0.024965525 & 0.003723482 & 0.017490262 & 0.009574812 & 0.004827486 & 0.005907706 & 0.010648910 & 0.004365652 & 0.012953572 & 0.008849781 & 0.010822306 & 0.005128003 & 0.007023774 & 0.010940720 & 0.032030102 & 0.013310011 & 0.008011489 & 0.004810598 & 0.014251505 & 0.002832420 & 0.02637555 & 0.005835421 & 0.012582770 & 0.007445882 & 0.016933151 & 0.005157431 & 0.006874123 \\\\\n\\end{tabular}\n","text/markdown":"\n| fname | Accelerating_and_revving_and_vroom | Accordion | Acoustic_guitar | Applause | Bark | Bass_drum | Bass_guitar | Bathtub_(filling_or_washing) | Bicycle_bell | ⋯ | Toilet_flush | Traffic_noise_and_roadway_noise | Trickle_and_dribble | Walk_and_footsteps | Water_tap_and_faucet | Waves_and_surf | Whispering | Writing | Yell | Zipper_(clothing) |\n|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| 000ccb97.wav | 0.013209159 | 0.01774904 | 0.002916300 | 0.009029981 | 0.019333581 | 0.009261215 | 0.01130432 | 0.003191851 | 0.041688547 | ⋯ | 0.021906769 | 0.004111980 | 0.011314391 | 0.01709291 | 0.005540514 | 0.001942767 | 0.013837142 | 0.003929277 | 0.002548119 | 0.002510201 |\n| 0012633b.wav | 0.017934894 | 0.01739276 | 0.008245957 | 0.005662751 | 0.019713445 | 0.018844701 | 0.02341015 | 0.005488900 | 0.009091829 | ⋯ | 0.007201934 | 0.013194967 | 0.005653424 | 0.01434080 | 0.012078281 | 0.013583345 | 0.006861777 | 0.010532438 | 0.009440523 | 0.007197834 |\n| 001ed5f1.wav | 0.014889697 | 0.01000734 | 0.014423068 | 0.012587036 | 0.008792087 | 0.010042071 | 0.01154019 | 0.010289351 | 0.017354663 | ⋯ | 0.015117173 | 0.010329596 | 0.011478883 | 0.01193406 | 0.008192242 | 0.012718624 | 0.009315304 | 0.012257267 | 0.015921116 | 0.016205560 |\n| 00294be0.wav | 0.023437565 | 0.01410034 | 0.012042203 | 0.005482092 | 0.023469465 | 0.023048503 | 0.03446521 | 0.004114904 | 0.007335012 | ⋯ | 0.004856232 | 0.016843693 | 0.004171766 | 0.01782716 | 0.008663573 | 0.019125421 | 0.007035829 | 0.014549707 | 0.007871445 | 0.009078503 |\n| 003fde7a.wav | 0.005750209 | 0.01126031 | 0.003639374 | 0.006456756 | 0.008438641 | 0.006241641 | 0.01137754 | 0.003585051 | 0.021493951 | ⋯ | 0.016109822 | 0.004421263 | 0.009133413 | 0.01445377 | 0.006500922 | 0.005644536 | 0.026017219 | 0.013965773 | 0.034930632 | 0.007944757 |\n| 0040ccc9.wav | 0.014303586 | 0.01032966 | 0.024132181 | 0.007142874 | 0.026701186 | 0.026633421 | 0.04539999 | 0.003292581 | 0.006904302 | ⋯ | 0.004810598 | 0.014251505 | 0.002832420 | 0.02637555 | 0.005835421 | 0.012582770 | 0.007445882 | 0.016933151 | 0.005157431 | 0.006874123 |\n\n","text/html":"\n\n\t\n\n\n\t\n\t\n\t\n\t\n\t\n\t\n\n
fnameAccelerating_and_revving_and_vroomAccordionAcoustic_guitarApplauseBarkBass_drumBass_guitarBathtub_(filling_or_washing)Bicycle_bell⋯Toilet_flushTraffic_noise_and_roadway_noiseTrickle_and_dribbleWalk_and_footstepsWater_tap_and_faucetWaves_and_surfWhisperingWritingYellZipper_(clothing)
000ccb97.wav0.013209159 0.01774904 0.002916300 0.009029981 0.019333581 0.009261215 0.01130432 0.003191851 0.041688547 ⋯ 0.021906769 0.004111980 0.011314391 0.01709291 0.005540514 0.001942767 0.013837142 0.003929277 0.002548119 0.002510201
0012633b.wav0.017934894 0.01739276 0.008245957 0.005662751 0.019713445 0.018844701 0.02341015 0.005488900 0.009091829 ⋯ 0.007201934 0.013194967 0.005653424 0.01434080 0.012078281 0.013583345 0.006861777 0.010532438 0.009440523 0.007197834
001ed5f1.wav0.014889697 0.01000734 0.014423068 0.012587036 0.008792087 0.010042071 0.01154019 0.010289351 0.017354663 ⋯ 0.015117173 0.010329596 0.011478883 0.01193406 0.008192242 0.012718624 0.009315304 0.012257267 0.015921116 0.016205560
00294be0.wav0.023437565 0.01410034 0.012042203 0.005482092 0.023469465 0.023048503 0.03446521 0.004114904 0.007335012 ⋯ 0.004856232 0.016843693 0.004171766 0.01782716 0.008663573 0.019125421 0.007035829 0.014549707 0.007871445 0.009078503
003fde7a.wav0.005750209 0.01126031 0.003639374 0.006456756 0.008438641 0.006241641 0.01137754 0.003585051 0.021493951 ⋯ 0.016109822 0.004421263 0.009133413 0.01445377 0.006500922 0.005644536 0.026017219 0.013965773 0.034930632 0.007944757
0040ccc9.wav0.014303586 0.01032966 0.024132181 0.007142874 0.026701186 0.026633421 0.04539999 0.003292581 0.006904302 ⋯ 0.004810598 0.014251505 0.002832420 0.02637555 0.005835421 0.012582770 0.007445882 0.016933151 0.005157431 0.006874123
\n"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"write.csv(predictions, 'submission.csv', row.names=FALSE)","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"display_name":"R","language":"R","name":"ir"},"language_info":{"mimetype":"text/x-r-source","name":"R","pygments_lexer":"r","version":"3.4.2","file_extension":".r","codemirror_mode":"r"}},"nbformat":4,"nbformat_minor":1}