Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

model with context 2 #31

Open
wants to merge 7 commits into
base: main
Choose a base branch
from
Open

model with context 2 #31

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
74b62a4
added a sanity check to compare stored OSM cube with df_agg (aggregat…
alkalait Jun 17, 2024
4479fe6
get_chip_ip: True
alkalait Jun 17, 2024
098549b
Dataset returns OSM cubes
alkalait Jun 17, 2024
0250ca8
test out OSM cubes; visualize
alkalait Jun 17, 2024
2c475be
fixed issue of binarizing and aggregating therefore creating >1 count…
alkalait Jun 18, 2024
a540a99
experiment with labels from 17 x17 grid
alkalait Jun 18, 2024
c9ac4ee
MultiscaleContextualCNN
alkalait Jun 20, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 3 additions & 1 deletion configs/naip-multilabel-contextual.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,10 +3,12 @@ dataloader:
metadata_file: /opt/data/california-naip-chips/california-naip-chips-100k.parquet
neighbor_embeddings_folder: /opt/data/california-naip-chips/california-naip-chips-100k-neighbours/npy
neighborhood_radius: 8 # max 8
collapse_osm_cube: True
get_osm_strlabels: True
get_osm_ohearea: True
get_osm_ohecount: True
get_osm_ohelength: True
get_chip_id: True
embeddings_normalization: True
osmvector_normalization: False
# multilabel_threshold_osm_ohecount: 1
Expand All @@ -16,7 +18,7 @@ dataloader:
num_workers: 4

model:
_target_: earthtext.models.multilabel.MultisizeContextualCNN
_target_: earthtext.models.multilabel.MultiscaleContextualCNN
input_dim: 768
output_dim: 140
layers_spec: [512, 256, 128]
Expand Down
314 changes: 179 additions & 135 deletions notebooks/models/04e - multilabel classification w context v1 NAIP.ipynb
View file
Open in desktop

Large diffs are not rendered by default.

148 changes: 125 additions & 23 deletions notebooks/naip/00c - neighbours 3D embedding arrays.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -198,15 +198,15 @@
},
{
"cell_type": "code",
"execution_count": null,
"execution_count": 9,
"id": "4f20d2d4-af19-49e6-86d2-94aa0d378171",
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"\u001b[38;2;255;117;0m 22%\u001b[39m \u001b[38;2;255;117;0m(23089 of 104234)\u001b[39m |### | Elapsed Time: 0:02:59 ETA: 0:10:32"
"\u001b[38;2;0;255;0m100%\u001b[39m \u001b[38;2;0;255;0m(104234 of 104234)\u001b[39m |################| Elapsed Time: 1:04:41 Time: 1:04:413110\n"
]
}
],
Expand Down Expand Up @@ -297,6 +297,7 @@
"source": [
"folder_grid = \"/opt/data/california-naip-chips/california-naip-chips-100k-neighbours/grid\"\n",
"folder_npy = \"/opt/data/california-naip-chips/california-naip-chips-100k-neighbours/npy\"\n",
"folder_osm = \"/opt/data/california-naip-chips/california-naip-chips-100k-neighbours/osm\"\n",
"file_grid = np.random.choice(os.listdir(folder_grid))\n",
"file_npy = np.random.choice(os.listdir(folder_npy))\n",
"df = pd.read_parquet(f\"{folder_grid}/{file_grid}\")\n",
Expand Down Expand Up @@ -336,18 +337,16 @@
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "78e15e82-98c9-4481-b5b7-d2cde0044eed",
"cell_type": "markdown",
"id": "0b6c922b-b3ad-4f1f-89b1-b3c60706fe98",
"metadata": {},
"outputs": [],
"source": [
"# TODO: Validate the OSM cube by checking it the sum over TAGS dimension equals the aggregated OSM"
"Validate the OSM cube by checking it the sum over TAGS dimension equals the aggregated OSM"
]
},
{
"cell_type": "code",
"execution_count": null,
"execution_count": 29,
"id": "5aa5601d-26bb-47ba-83f6-4f606aa114fe",
"metadata": {},
"outputs": [],
Expand All @@ -360,17 +359,7 @@
},
{
"cell_type": "code",
"execution_count": 5,
"id": "33822830-d4db-4991-b7a5-eb85cd46c83b",
"metadata": {},
"outputs": [],
"source": [
"# df_agg = pd.read_parquet(f'{folder}/osm_aggregate.parquet')"
]
},
{
"cell_type": "code",
"execution_count": 7,
"execution_count": 30,
"id": "3727a4d0-fc79-4438-a173-59f609db2404",
"metadata": {
"scrolled": true
Expand All @@ -380,12 +369,14 @@
"name": "stderr",
"output_type": "stream",
"text": [
"\u001b[38;2;0;255;0m100%\u001b[39m \u001b[38;2;0;255;0m(3210 of 3210)\u001b[39m |####################| Elapsed Time: 0:01:39 Time: 0:01:390008\n"
"\u001b[38;2;0;255;0m100%\u001b[39m \u001b[38;2;0;255;0m(1 of 1)\u001b[39m |##########################| Elapsed Time: 0:00:00 Time: 0:00:00\n"
]
}
],
"source": [
"for i in pbar(range(len(df_agg))):\n",
"# CAUTION before you run: takes an hour\n",
"# for i in pbar(range(len(df_agg))):\n",
"for i in pbar(range(1)):\n",
" row = df_agg.iloc[i]\n",
" z = pd.read_parquet(f\"{folder}/{row.name}.parquet\")\n",
" _osm = _df.loc[z.chipid[z.chipid.isin(_df.index)]].sum()\n",
Expand All @@ -394,12 +385,123 @@
},
{
"cell_type": "code",
"execution_count": 8,
"execution_count": 31,
"id": "6bf25a90-ba76-4f54-9d4e-4134112b97c9",
"metadata": {},
"outputs": [],
"source": [
"df_agg.to_parquet(f'{folder}/osm_aggregate.parquet')"
"## CAUTION: will overwrite the existing .parquet file\n",
"# df_agg.to_parquet(f'{folder}/osm_aggregate.parquet')"
]
},
{
"cell_type": "code",
"execution_count": 33,
"id": "9eec7ac2-7fc1-4510-94fa-9ab74bfe6533",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"ca_m_3411730_sw_11_060_20220501-36-11\n"
]
},
{
"data": {
"text/plain": [
"onehot_count [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 13, 0,...\n",
"onehot_area [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...\n",
"onehot_length [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...\n",
"Name: ca_m_3411730_sw_11_060_20220501-36-11, dtype: object"
]
},
"execution_count": 33,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"print(row.name)\n",
"df_agg.loc[row.name]"
]
},
{
"cell_type": "code",
"execution_count": 39,
"id": "d0ddd3b8-6ca4-4680-9ed6-6d925189f325",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(140, 3, 17, 17)"
]
},
"execution_count": 39,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"x = np.load(f'{folder}/osm/{row.name}.osm.npy')\n",
"x.shape"
]
},
{
"cell_type": "markdown",
"id": "fdb4d999-8144-4821-8176-1eb173cdc4af",
"metadata": {},
"source": [
"Compare. NOTE: `df_agg` aggregates osm data from ALL neighbors in the circle, beyond the 17x17 grid.\n",
"\n",
"But they look similar enough for the purposes of this sanity check."
]
},
{
"cell_type": "code",
"execution_count": 50,
"id": "c257178e-4afb-4f77-b537-13d54f507503",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[ 0 0 0 0 0 0 0 0 0 0 0 0 0 13 0 0 0 0 0 5 0 0 0 0\n",
" 0 8 0 0 0 0 0 0 0 0 56 0 0 0 0 49 7 0 0 28 0 0 4 0\n",
" 0 0 0 0 2 0 0 0 0 0 0 0 0 22 0 0 0 0 0 0 0 0 0 0\n",
" 15 0 0 0 0 12 0 0 0 0 3 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
" 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
" 0 0 0 0 0 0 0 0 0 0 0 0 0 24 0 24 9 0 3 6]\n"
]
}
],
"source": [
"print(df_agg.loc[row.name]['onehot_count'])"
]
},
{
"cell_type": "code",
"execution_count": 59,
"id": "d696c4b2-0edd-417a-8fd9-282f307ca05c",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[ 0 0 0 0 0 0 0 0 0 0 0 0 0 8 0 0 0 0 0 3 0 0 0 0\n",
" 0 5 0 0 0 0 0 0 0 0 25 0 0 0 0 23 2 0 0 14 0 0 4 0\n",
" 0 0 0 0 1 0 0 0 0 0 0 0 0 9 0 0 0 0 0 0 0 0 0 0\n",
" 8 0 0 0 0 8 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
" 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
" 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 0 0 3]\n"
]
}
],
"source": [
"print(x.sum(axis=(2,3))[:, 0].astype(int))"
]
}
],
Expand Down
Loading