|
| 1 | +<!-- |
| 2 | +
|
| 3 | + Licensed to the Apache Software Foundation (ASF) under one |
| 4 | + or more contributor license agreements. See the NOTICE file |
| 5 | + distributed with this work for additional information |
| 6 | + regarding copyright ownership. The ASF licenses this file |
| 7 | + to you under the Apache License, Version 2.0 (the |
| 8 | + "License"); you may not use this file except in compliance |
| 9 | + with the License. You may obtain a copy of the License at |
| 10 | + |
| 11 | + http://www.apache.org/licenses/LICENSE-2.0 |
| 12 | + |
| 13 | + Unless required by applicable law or agreed to in writing, |
| 14 | + software distributed under the License is distributed on an |
| 15 | + "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY |
| 16 | + KIND, either express or implied. See the License for the |
| 17 | + specific language governing permissions and limitations |
| 18 | + under the License. |
| 19 | +
|
| 20 | +--> |
| 21 | +# Set Operations |
| 22 | + |
| 23 | +IoTDB natively supports standard SQL set operations, including three core operators: **UNION**, **INTERSECT**, and **EXCEPT**. These operations enable seamless merging, comparison, and filtering of query results from multiple time-series data sources, greatly improving the flexibility and efficiency of time-series data analysis. |
| 24 | + |
| 25 | +> Note: This feature is available since version 2.0.9-beta. |
| 26 | +
|
| 27 | +## 1. UNION |
| 28 | +### 1.1 Overview |
| 29 | +The UNION operator combines all rows from two result sets (order not guaranteed), supporting both duplicate elimination (default) and duplicate retention modes. |
| 30 | + |
| 31 | +### 1.2 Syntax |
| 32 | +```sql |
| 33 | +query UNION (ALL | DISTINCT) query |
| 34 | +``` |
| 35 | + |
| 36 | +**Description** |
| 37 | +1. **Duplicate Handling** |
| 38 | + - Default (`UNION` or `UNION DISTINCT`): Automatically removes duplicate rows. |
| 39 | + - `UNION ALL`: Preserves all rows (including duplicates) with higher performance. |
| 40 | + |
| 41 | +2. **Input Requirements** |
| 42 | + - The two queries must return the same number of columns. |
| 43 | + - Corresponding columns must have compatible data types: |
| 44 | + - Numeric compatibility: `INT32`, `INT64`, `FLOAT`, and `DOUBLE` are fully compatible with each other. |
| 45 | + - String compatibility: `TEXT` and `STRING` are fully compatible. |
| 46 | + - Special rule: `INT64` is compatible with `TIMESTAMP`. |
| 47 | + |
| 48 | +3. **Result Set Rules** |
| 49 | + - Column names and order are inherited from the first query. |
| 50 | + |
| 51 | +### 1.3 Examples |
| 52 | +Using the [sample data](../Reference/Sample-Data.md): |
| 53 | + |
| 54 | +1. Get distinct non-null device and temperature records from `table1` and `table2` |
| 55 | +```sql |
| 56 | +SELECT device_id, temperature FROM table1 WHERE temperature IS NOT NULL |
| 57 | +UNION |
| 58 | +SELECT device_id, temperature FROM table2 WHERE temperature IS NOT NULL; |
| 59 | + |
| 60 | +-- Equivalent to: |
| 61 | +SELECT device_id, temperature FROM table1 WHERE temperature IS NOT NULL |
| 62 | +UNION DISTINCT |
| 63 | +SELECT device_id, temperature FROM table2 WHERE temperature IS NOT NULL; |
| 64 | +``` |
| 65 | + |
| 66 | +Result: |
| 67 | +``` |
| 68 | ++---------+-----------+ |
| 69 | +|device_id|temperature| |
| 70 | ++---------+-----------+ |
| 71 | +| 101| 90.0| |
| 72 | +| 101| 85.0| |
| 73 | +| 100| 90.0| |
| 74 | +| 100| 85.0| |
| 75 | +| 100| 88.0| |
| 76 | ++---------+-----------+ |
| 77 | +Total line number = 5 |
| 78 | +It costs 0.074s |
| 79 | +``` |
| 80 | + |
| 81 | +2. Get all non-null device and temperature records from `table1` and `table2` (including duplicates) |
| 82 | +```sql |
| 83 | +SELECT device_id, temperature FROM table1 WHERE temperature IS NOT NULL |
| 84 | +UNION ALL |
| 85 | +SELECT device_id, temperature FROM table2 WHERE temperature IS NOT NULL; |
| 86 | +``` |
| 87 | + |
| 88 | +Result: |
| 89 | +``` |
| 90 | ++---------+-----------+ |
| 91 | +|device_id|temperature| |
| 92 | ++---------+-----------+ |
| 93 | +| 101| 90.0| |
| 94 | +| 101| 90.0| |
| 95 | +| 101| 85.0| |
| 96 | +| 101| 85.0| |
| 97 | +| 101| 85.0| |
| 98 | +| 101| 85.0| |
| 99 | +| 100| 90.0| |
| 100 | +| 100| 85.0| |
| 101 | +| 100| 85.0| |
| 102 | +| 100| 88.0| |
| 103 | +| 100| 90.0| |
| 104 | +| 100| 90.0| |
| 105 | +| 101| 90.0| |
| 106 | +| 101| 85.0| |
| 107 | +| 101| 85.0| |
| 108 | +| 100| 85.0| |
| 109 | +| 100| 90.0| |
| 110 | ++---------+-----------+ |
| 111 | +Total line number = 17 |
| 112 | +It costs 0.108s |
| 113 | +``` |
| 114 | + |
| 115 | +> **Notes** |
| 116 | +> - Set operations **do not guarantee result order**; actual output may differ from examples. |
| 117 | +
|
| 118 | + |
| 119 | +## 2. INTERSECT |
| 120 | +### 2.1 Overview |
| 121 | +The INTERSECT operator returns rows that exist in both result sets (order not guaranteed), supporting both duplicate elimination (default) and duplicate retention modes. |
| 122 | + |
| 123 | +### 2.2 Syntax |
| 124 | +```sql |
| 125 | +query1 INTERSECT [ALL | DISTINCT] query2 |
| 126 | +``` |
| 127 | + |
| 128 | +**Description** |
| 129 | +1. **Duplicate Handling** |
| 130 | + - Default (`INTERSECT` or `INTERSECT DISTINCT`): Automatically removes duplicate rows. |
| 131 | + - `INTERSECT ALL`: Preserves duplicate rows, with slightly lower performance. |
| 132 | + |
| 133 | +2. **Precedence Rules** |
| 134 | + - `INTERSECT` has higher precedence than `UNION` and `EXCEPT` |
| 135 | + (e.g., `A UNION B INTERSECT C` is equivalent to `A UNION (B INTERSECT C)`). |
| 136 | + - Evaluation is left-to-right |
| 137 | + (e.g., `A INTERSECT B INTERSECT C` is equivalent to `(A INTERSECT B) INTERSECT C`). |
| 138 | + |
| 139 | +3. **Input Requirements** |
| 140 | + - The two queries must return the same number of columns. |
| 141 | + - Corresponding columns must have compatible data types (same rules as UNION). |
| 142 | + - NULL values are treated as equal (`NULL IS NOT DISTINCT FROM NULL`). |
| 143 | + - If the `time` column is not included in `SELECT`, it does not participate in comparison and will not appear in the result. |
| 144 | + |
| 145 | +4. **Result Set Rules** |
| 146 | + - Column names and order are inherited from the first query. |
| 147 | + |
| 148 | +### 2.3 Examples |
| 149 | +Using the [sample data](../Reference/Sample-Data.md): |
| 150 | + |
| 151 | +1. Get distinct common device and temperature records from `table1` and `table2` |
| 152 | +```sql |
| 153 | +SELECT device_id, temperature FROM table1 |
| 154 | +INTERSECT |
| 155 | +SELECT device_id, temperature FROM table2; |
| 156 | + |
| 157 | +-- Equivalent to: |
| 158 | +SELECT device_id, temperature FROM table1 |
| 159 | +INTERSECT DISTINCT |
| 160 | +SELECT device_id, temperature FROM table2; |
| 161 | +``` |
| 162 | + |
| 163 | +Result: |
| 164 | +``` |
| 165 | ++---------+-----------+ |
| 166 | +|device_id|temperature| |
| 167 | ++---------+-----------+ |
| 168 | +| 101| 90.0| |
| 169 | +| 101| 85.0| |
| 170 | +| 100| null| |
| 171 | +| 100| 90.0| |
| 172 | +| 100| 85.0| |
| 173 | ++---------+-----------+ |
| 174 | +Total line number = 5 |
| 175 | +It costs 0.087s |
| 176 | +``` |
| 177 | + |
| 178 | +2. Get all common device and temperature records from `table1` and `table2` (including duplicates) |
| 179 | +```sql |
| 180 | +SELECT device_id, temperature FROM table1 |
| 181 | +INTERSECT ALL |
| 182 | +SELECT device_id, temperature FROM table2; |
| 183 | +``` |
| 184 | + |
| 185 | +Result: |
| 186 | +``` |
| 187 | ++---------+-----------+ |
| 188 | +|device_id|temperature| |
| 189 | ++---------+-----------+ |
| 190 | +| 100| 85.0| |
| 191 | +| 100| 90.0| |
| 192 | +| 100| null| |
| 193 | +| 101| 85.0| |
| 194 | +| 101| 85.0| |
| 195 | +| 101| 90.0| |
| 196 | ++---------+-----------+ |
| 197 | +Total line number = 6 |
| 198 | +It costs 0.139s |
| 199 | +``` |
| 200 | + |
| 201 | +> **Notes** |
| 202 | +> - Set operations **do not guarantee result order**. |
| 203 | +> - When mixed with `UNION`/`EXCEPT`, use parentheses to explicitly specify precedence |
| 204 | + > (e.g., `A INTERSECT (B UNION C)`). |
| 205 | +
|
| 206 | + |
| 207 | +## 3. EXCEPT |
| 208 | +### 3.1 Overview |
| 209 | +The EXCEPT operator returns rows that exist in the first result set but **not** in the second (order not guaranteed), supporting both duplicate elimination (default) and duplicate retention modes. |
| 210 | + |
| 211 | +### 3.2 Syntax |
| 212 | +```sql |
| 213 | +query1 EXCEPT [ALL | DISTINCT] query2 |
| 214 | +``` |
| 215 | + |
| 216 | +**Description** |
| 217 | +1. **Duplicate Handling** |
| 218 | + - Default (`EXCEPT` or `EXCEPT DISTINCT`): Automatically removes duplicate rows. |
| 219 | + - `EXCEPT ALL`: Preserves duplicate rows, with slightly lower performance. |
| 220 | + |
| 221 | +2. **Precedence Rules** |
| 222 | + - `EXCEPT` has the same precedence as `UNION`, and lower precedence than `INTERSECT` |
| 223 | + (e.g., `A INTERSECT B EXCEPT C` is equivalent to `(A INTERSECT B) EXCEPT C`). |
| 224 | + - Evaluation is left-to-right |
| 225 | + (e.g., `A EXCEPT B EXCEPT C` is equivalent to `(A EXCEPT B) EXCEPT C`). |
| 226 | + |
| 227 | +3. **Input Requirements** |
| 228 | + - The two queries must return the same number of columns. |
| 229 | + - Corresponding columns must have compatible data types (same rules as UNION). |
| 230 | + - NULL values are treated as equal (`NULL IS NOT DISTINCT FROM NULL`). |
| 231 | + - If the `time` column is not included in `SELECT`, it does not participate in comparison and will not appear in the result. |
| 232 | + |
| 233 | +4. **Result Set Rules** |
| 234 | + - Column names and order are inherited from the first query. |
| 235 | + |
| 236 | +### 3.3 Examples |
| 237 | +Using the [sample data](../Reference/Sample-Data.md): |
| 238 | + |
| 239 | +1. Get distinct records from `table1` that do not exist in `table2` |
| 240 | +```sql |
| 241 | +SELECT device_id, temperature FROM table1 |
| 242 | +EXCEPT |
| 243 | +SELECT device_id, temperature FROM table2; |
| 244 | + |
| 245 | +-- Equivalent to: |
| 246 | +SELECT device_id, temperature FROM table1 |
| 247 | +EXCEPT DISTINCT |
| 248 | +SELECT device_id, temperature FROM table2; |
| 249 | +``` |
| 250 | + |
| 251 | +Result: |
| 252 | +``` |
| 253 | ++---------+-----------+ |
| 254 | +|device_id|temperature| |
| 255 | ++---------+-----------+ |
| 256 | +| 101| null| |
| 257 | +| 100| 88.0| |
| 258 | ++---------+-----------+ |
| 259 | +Total line number = 2 |
| 260 | +It costs 0.173s |
| 261 | +``` |
| 262 | + |
| 263 | +2. Get all records from `table1` that do not exist in `table2` (including duplicates) |
| 264 | +```sql |
| 265 | +SELECT device_id, temperature FROM table1 |
| 266 | +EXCEPT ALL |
| 267 | +SELECT device_id, temperature FROM table2; |
| 268 | +``` |
| 269 | + |
| 270 | +Result: |
| 271 | +``` |
| 272 | ++---------+-----------+ |
| 273 | +|device_id|temperature| |
| 274 | ++---------+-----------+ |
| 275 | +| 100| 85.0| |
| 276 | +| 100| 88.0| |
| 277 | +| 100| 90.0| |
| 278 | +| 100| 90.0| |
| 279 | +| 100| null| |
| 280 | +| 101| 85.0| |
| 281 | +| 101| 85.0| |
| 282 | +| 101| 90.0| |
| 283 | +| 101| null| |
| 284 | +| 101| null| |
| 285 | +| 101| null| |
| 286 | +| 101| null| |
| 287 | ++---------+-----------+ |
| 288 | +Total line number = 12 |
| 289 | +It costs 0.155s |
| 290 | +``` |
| 291 | + |
| 292 | +> **Notes** |
| 293 | +> - Set operations **do not guarantee result order**. |
| 294 | +> - When mixed with `UNION`/`INTERSECT`, use parentheses to explicitly specify precedence |
| 295 | + > (e.g., `A EXCEPT (B INTERSECT C)`). |
0 commit comments