Browse Source

No, 결측이상치처리 should be inside the 이력+이동류. So the order is as before, except that processing of outlier and missing value should be taken place on rhistory, not hrhistory or so.

master
김선중 1 year ago
parent
commit
ad520b053a
4 changed files with 2775 additions and 9498 deletions
  1. +957
    -73
      Analysis/0109_movement_matching/0111_preprocess_5.ipynb
  2. +1816
    -9423
      Analysis/0109_movement_matching/0111_preprocess_6.ipynb
  3. +2
    -2
      Data/tables/raw_tables/0110_make_sample_table.ipynb
  4. BIN
      Documents/1127_table_definition/table_definition_v0.8.3.xlsx

+ 957
- 73
Analysis/0109_movement_matching/0111_preprocess_5.ipynb
File diff suppressed because it is too large
View File


+ 1816
- 9423
Analysis/0109_movement_matching/0111_preprocess_6.ipynb
File diff suppressed because it is too large
View File


+ 2
- 2
Data/tables/raw_tables/0110_make_sample_table.ipynb View File

@ -1136,7 +1136,7 @@
},
{
"cell_type": "code",
"execution_count": 20,
"execution_count": 21,
"metadata": {},
"outputs": [
{
@ -2890,6 +2890,7 @@
" plan[col] = 0\n",
"plan.reset_index(inplace=True)\n",
"plan = plan[list(plan.columns)[:3] + sorted(plan.columns)[1:-4] + list(plan.columns)[3:5]]\n",
"plan.to_csv('../plan.csv')\n",
"pland = plan.copy()\n",
"cums_A = np.array(pland[[f'dura_A{k}' for k in range(1,9)]].cumsum(axis=1))\n",
"cums_B = np.array(pland[[f'dura_B{k}' for k in range(1,9)]].cumsum(axis=1))\n",
@ -2905,7 +2906,6 @@
" i+=1\n",
"pland = pland[['inter_no', 'start_hour', 'start_minute'] + [f'ddur_{i}' for i in range(1,9)] + ['cycle', 'offset']] # plan detailed\n",
"pland[[f'ddur_{i}' for i in range(1,9)]] = pland[[f'ddur_{i}' for i in range(1,9)]].astype(int)\n",
"pland.to_csv('../plan.csv')\n",
"display(plan)\n",
"display(pland)"
]

BIN
Documents/1127_table_definition/table_definition_v0.8.3.xlsx View File


Loading…
Cancel
Save