feat(chore): Change train_test to ghout dir, add ghout to .gitignore'

.gitignore

@@ -131,6 +131,7 @@ dmypy.json
 # Honeybee/Ladybug data
 /run
 /models
+/ghout
 *.sql
 
 

deeprad/image2vec.py

@@ -4,7 +4,6 @@
 import os
 import json
 from copy import copy
-from pathlib import Path
 
 # scientific python stuff
 import cv2
@@ -16,9 +15,9 @@
 from sklearn.mixture import GaussianMixture
 from matplotlib.axes import Axes
 
-# Path to all models in deep_rad
-DEEPRAD_DATA_DIR = os.path.abspath(os.path.join(
-    os.getcwd(), '..', 'deeprad/data/models/'))
+# deeprad modules
+from .utils import to_poly_sh, to_poly_np, extract_floorplan_ids, load_floorplan_data, \
+    load_json, make_dir_safely, DEEPRAD_MODELS_DIR
 
 
 def viz_poly_sh_arr(poly_sh_arr, a=None, scale=1, iter=False, **kwargs):
@@ -63,28 +62,6 @@ def viz_ortho_snap_grid(x_means: np.ndarray, y_means: np.ndarray, ax: Axes, colo
     return ax
 
 
-def to_poly_sh(xy_arr):
-    """Shapely polygon from list of two arrays of x, y coordinates.
-
-        Args:
-            xy_arr: [x_arr, y_arr]
-    Example:
-        to_poly_sh(to_poly_np(poly_sh))
-        to_poly_np(to_poly_sh(poly_np))
-    """
-    return geom.Polygon([(x, y) for x, y in zip(*xy_arr)])
-
-
-def to_poly_np(poly_sh):
-    """Two arrays of x, y coordinates from shapely polygon.
-
-    Example:
-        to_poly_sh(to_poly_np(poly_sh))
-        to_poly_np(to_poly_sh(poly_np))
-    """
-    return np.array(poly_sh.exterior.xy)
-
-
 def rooms_equal_color_mapping(_rooms: np.ndarray) -> np.ndarray:
     """Replace class indices with equally spaced intervals from 0-255 for thresholding.
 
@@ -420,79 +397,6 @@ def _get_cluster_coord_fx(polyc):
     return _get_cluster_coord_fx
 
 
-def extract_floorplan_ids(data_num):
-    """Safely extract root model directories for polygon extraction."""
-
-    # Load all model directories
-    floorplan_id_arr = os.listdir(DEEPRAD_DATA_DIR)
-    n_ids = len(floorplan_id_arr)
-
-    if n_ids == 0:
-        raise Exception('No image files. Add images and try again.')
-
-    if data_num > n_ids:
-        print('Note: data_num {} is too high. Resetting to n_ids {}.'.format(
-            data_num, n_ids))
-        data_num = n_ids
-
-    return data_num, floorplan_id_arr
-
-
-def load_json(json_fpath):
-    with open(json_fpath, 'r') as fp:
-        val_dict = json.load(fp)
-
-    return val_dict
-
-
-def load_floorplan_data(targ_id_dirs, data_num):
-    """Load floorplan data."""
-
-    src_img_arr = [0] * data_num
-    label_img_arr = [0] * data_num
-    hdict_arr = [0] * data_num
-    targ_id_dir_arr = [0] * data_num
-
-    i = -1
-    idx = -1
-    total_i = 0
-    null_lst = []
-
-    while (idx + 1) < data_num:
-        i += 1
-        targ_id_dir = os.path.join(
-            DEEPRAD_DATA_DIR, targ_id_dirs[i], 'data')
-        targ_src_fpath = os.path.join(targ_id_dir, 'src.jpg')
-        targ_label_fpath = os.path.join(targ_id_dir, 'label.jpg')
-        targ_json_fpath = os.path.join(targ_id_dir, 'door_vecs.json')
-
-        hdict = load_json(targ_json_fpath)
-
-        if hdict['scale'] < 0.4:
-            print('Skip {} b/c scale at {}. Total skipped={}.'.format(
-                  targ_id_dirs[i], hdict['scale'], total_i))
-            total_i += 1
-            continue
-
-        idx += 1
-        hdict_arr[idx] = hdict
-        src_img_arr[idx] = cv2.imread(targ_src_fpath, cv2.COLOR_BGR2RGB)
-        label_img_arr[idx] = cv2.imread(targ_label_fpath, cv2.IMREAD_GRAYSCALE)
-        targ_id_dir_arr[idx] = targ_id_dir
-        null_lst.append(targ_id_dirs[i] + '\n')
-
-    # Write to null list
-    null_fpath = os.path.join(DEEPRAD_DATA_DIR, '_null.txt')
-    with open(null_fpath, 'w') as fp:
-        #[fp.writeline(null_) for null_ in null_lst]
-        fp.writelines(null_lst)
-    return hdict_arr, src_img_arr, label_img_arr, targ_id_dir_arr
-
-
-def make_dir_safely(dest_dir):
-    Path(dest_dir).mkdir(parents=True, exist_ok=True)
-
-
 def main(data_num):
     """Generate data_num amount of floorplan polygon vectors."""
 
@@ -617,7 +521,6 @@ def _clean_poly_sh(p): return clean_poly_sh(p, beps, stol, dpct)
         # -------------------------------------------------------------
         # Compute gaussian mixture models
         # -------------------------------------------------------------
-
         random_seed = 111
         poly_sh_arr = [copy(p) for p in _poly_sh_arr]
 

deeprad/preprocess_traintest.py

@@ -0,0 +1,51 @@
+"""Preprocess raw outputs from GH for train/test dataset."""
+import os
+import numpy as np
+
+import cv2
+
+import deeprad.utils as utils
+import deeprad.image2vec as i2v
+
+DEEPRAD_GHOUT_DIR = utils.DEEPRAD_GHOUT_DIR
+DEEPRAD_TRAINTEST_DIR = utils.DEEPRAD_TRAINTEST_DIR
+
+
+def load_img_gray(img_fpath):
+    return cv2.imread(img_fpath, cv2.)
+
+
+def preprocess_img(img_fpath):
+    pass
+
+
+def main():
+    """Preproces main."""
+
+    # load all directories in train/test
+    data_num, floorplan_ids = \
+        utils.extract_floorplan_ids(1e6, DEEPRAD_TRAINTEST_DIR, verbose=False)
+
+    # for i, floorplan_id in enumerate(floorplan_ids):
+    i = 0
+    floorplan_id = floorplan_ids[0]
+
+    in_dir = os.path.join(DEEPRAD_GHOUT_DIR, floorplan_id, 'in_data')
+    out_dir = os.path.join(DEEPRAD_GHOUT_DIR, floorplan_id, 'out_label')
+
+    # Preprocess labels
+    label_fpaths = [os.path.join(out_dir, out) for out in os.listdir(out_dir)]
+    label_imgs = [preprocess_img(fpath) for fpath in label_fpaths]
+    #[save_img(img) for img in label_imgs]
+
+    # Preprocess input images
+    # add as channels
+
+    # save new images
+    # delete all raw images
+
+    pass
+
+
+if __name__ == "__main__":
+    main()

deeprad/utils.py

@@ -0,0 +1,114 @@
+"""Utility functions for deeprad."""
+import sys
+import os
+import json
+from pathlib import Path
+import matplotlib.pyplot as plt
+import numpy as np
+
+# Path to all models in deep_rad
+DEEPRAD_MODELS_DIR = os.path.abspath(os.path.join(
+    os.getcwd(), '..', 'deeprad/data/models/'))
+DEEPRAD_GHOUT_DIR = os.path.abspath(os.path.join(
+    os.getcwd(), '..', 'deeprad/data/ghout/'))
+DEEPRAD_TRAINTEST_DIR = os.path.abspath(os.path.join(
+    os.getcwd(), '..', 'deeprad/data/traintest/'))
+
+
+def to_poly_sh(xy_arr):
+    """Shapely polygon from list of two arrays of x, y coordinates.
+
+        Args:
+            xy_arr: [x_arr, y_arr]
+    Example:
+        to_poly_sh(to_poly_np(poly_sh))
+        to_poly_np(to_poly_sh(poly_np))
+    """
+    return geom.Polygon([(x, y) for x, y in zip(*xy_arr)])
+
+
+def to_poly_np(poly_sh):
+    """Two arrays of x, y coordinates from shapely polygon.
+
+    Example:
+        to_poly_sh(to_poly_np(poly_sh))
+        to_poly_np(to_poly_sh(poly_np))
+    """
+    return np.array(poly_sh.exterior.xy)
+
+
+def extract_floorplan_ids(data_num, target_data_dir=None, verbose=True):
+    """Safely extract root model directories for polygon extraction."""
+
+    if target_data_dir is None:
+        target_data_dir = DEEPRAD_MODELS_DIR
+
+    # Load all model directories
+    floorplan_id_arr = os.listdir(target_data_dir)
+    n_ids = len(floorplan_id_arr)
+
+    if n_ids == 0:
+        raise Exception('No image files. Add images and try again.')
+
+    if data_num > n_ids:
+        if verbose:
+            print('Note: data_num {} is too high. Resetting to n_ids {}.'.format(
+                  data_num, n_ids))
+        data_num = n_ids
+
+    return data_num, floorplan_id_arr
+
+
+def load_json(json_fpath):
+    with open(json_fpath, 'r') as fp:
+        val_dict = json.load(fp)
+
+    return val_dict
+
+
+def load_floorplan_data(targ_id_dirs, data_num):
+    """Load floorplan data."""
+
+    src_img_arr = [0] * data_num
+    label_img_arr = [0] * data_num
+    hdict_arr = [0] * data_num
+    targ_id_dir_arr = [0] * data_num
+
+    i = -1
+    idx = -1
+    total_i = 0
+    null_lst = []
+
+    while (idx + 1) < data_num:
+        i += 1
+        targ_id_dir = os.path.join(
+            DEEPRAD_MODELS_DIR, targ_id_dirs[i], 'data')
+        targ_src_fpath = os.path.join(targ_id_dir, 'src.jpg')
+        targ_label_fpath = os.path.join(targ_id_dir, 'label.jpg')
+        targ_json_fpath = os.path.join(targ_id_dir, 'door_vecs.json')
+
+        hdict = load_json(targ_json_fpath)
+
+        if hdict['scale'] < 0.4:
+            print('Skip {} b/c scale at {}. Total skipped={}.'.format(
+                  targ_id_dirs[i], hdict['scale'], total_i))
+            total_i += 1
+            continue
+
+        idx += 1
+        hdict_arr[idx] = hdict
+        src_img_arr[idx] = cv2.imread(targ_src_fpath, cv2.COLOR_BGR2RGB)
+        label_img_arr[idx] = cv2.imread(targ_label_fpath, cv2.IMREAD_GRAYSCALE)
+        targ_id_dir_arr[idx] = targ_id_dir
+        null_lst.append(targ_id_dirs[i] + '\n')
+
+    # Write to null list
+    null_fpath = os.path.join(DEEPRAD_MODELS_DIR, '_null.txt')
+    with open(null_fpath, 'w') as fp:
+        #[fp.writeline(null_) for null_ in null_lst]
+        fp.writelines(null_lst)
+    return hdict_arr, src_img_arr, label_img_arr, targ_id_dir_arr
+
+
+def make_dir_safely(dest_dir):
+    Path(dest_dir).mkdir(parents=True, exist_ok=True)