zsplit 1.0.1__tar.gz → 1.0.2__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (12) hide show
  1. {zsplit-1.0.1/zsplit.egg-info → zsplit-1.0.2}/PKG-INFO +39 -20
  2. {zsplit-1.0.1 → zsplit-1.0.2}/README.md +38 -19
  3. {zsplit-1.0.1 → zsplit-1.0.2}/pyproject.toml +1 -1
  4. {zsplit-1.0.1 → zsplit-1.0.2/zsplit.egg-info}/PKG-INFO +39 -20
  5. {zsplit-1.0.1 → zsplit-1.0.2}/LICENSE +0 -0
  6. {zsplit-1.0.1 → zsplit-1.0.2}/setup.cfg +0 -0
  7. {zsplit-1.0.1 → zsplit-1.0.2}/zsplit/__init__.py +0 -0
  8. {zsplit-1.0.1 → zsplit-1.0.2}/zsplit/core.py +0 -0
  9. {zsplit-1.0.1 → zsplit-1.0.2}/zsplit.egg-info/SOURCES.txt +0 -0
  10. {zsplit-1.0.1 → zsplit-1.0.2}/zsplit.egg-info/dependency_links.txt +0 -0
  11. {zsplit-1.0.1 → zsplit-1.0.2}/zsplit.egg-info/requires.txt +0 -0
  12. {zsplit-1.0.1 → zsplit-1.0.2}/zsplit.egg-info/top_level.txt +0 -0
{zsplit-1.0.1/zsplit.egg-info → zsplit-1.0.2}/PKG-INFO RENAMED
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: zsplit
3
- Version: 1.0.1
3
+ Version: 1.0.2
4
4
  Summary: Z-Split Normalization: Zero-Preserving Split Normalization for bipolar spectral indices
5
5
  Author-email: Abdulrhman Almoadi <aalmoadi@kacst.edu.sa>
6
6
  License: MIT License
@@ -58,7 +58,7 @@ Dynamic: license-file
58
58
 
59
59
  # Z-Split Normalization (Zero-Preserving Split Normalization)
60
60
 
61
- **Version:** 1.0.0
61
+ **Version:** 1.0.1
62
62
  **License:** MIT
63
63
 
64
64
  ---
@@ -132,6 +132,24 @@ result = normalize(your_bipolar_index)
132
132
 
133
133
  ---
134
134
 
135
+ ## Practical Applications in Remote Sensing
136
+
137
+ Z-Split addresses a critical problem in any workflow involving bipolar indices across time or sensors:
138
+
139
+ **1. Change Detection**
140
+ Applying Min-Max before change detection introduces artificial trends that appear as land cover change but reflect only normalization artifacts. A report showing "15% vegetation decline" may be entirely attributable to normalization, not reality.
141
+
142
+ **2. Time Series Analysis**
143
+ Studies tracking floods, drought, or urban expansion over time require a clean temporal signal. Min-Max corrupts this signal by rescaling each scene independently to [0, 1], making year-to-year comparisons unreliable.
144
+
145
+ **3. Multi-temporal Machine Learning**
146
+ Models trained on Min-Max normalized multi-year data learn spurious patterns introduced by normalization rather than real spectral change. Z-Split ensures the training signal reflects actual surface conditions.
147
+
148
+ **4. InSAR and Displacement Monitoring**
149
+ In LOS displacement data, zero means no movement. Shifting zero through Min-Max misclassifies stable pixels as deforming, directly corrupting subsidence or uplift maps.
150
+
151
+ ---
152
+
135
153
  ## Validation
136
154
 
137
155
  ### 1. NDWI — Optical Remote Sensing
@@ -142,7 +160,20 @@ Tested on Sentinel-2 NDWI data. Z-Split successfully expanded near-zero clustere
142
160
 
143
161
  Validated on four Sentinel-2 NDVI scenes (March–April, 2018–2020–2022–2024) over Riyadh, Saudi Arabia (scale: 10 m, 5603 × 5129 pixels per scene) across 11,424 spatial patches.
144
162
 
145
- #### A. Linear Trend (Slope per pixel, 2018–2024)
163
+ #### A. Normalization Artifact Analysis
164
+
165
+ The figure below shows the deviation of each normalization method from the raw temporal signal. Min-Max introduces large artificial trends across the entire scene (colored map). Z-Split introduces virtually no deviation (near-white map).
166
+
167
+ ![Deviation Analysis](https://raw.githubusercontent.com/aalmoadi/endwi/main/Figures/deviation_analysis.png)
168
+
169
+ | Method | Mean Absolute Deviation | Std |
170
+ |--------|:-----------------------:|:---:|
171
+ | Min-Max | 0.1858 | 0.2091 |
172
+ | **Z-Split** | **0.0013** | **0.0017** |
173
+
174
+ Z-Split preserves the temporal signal with **143× greater fidelity** than Min-Max.
175
+
176
+ #### B. Linear Trend (Slope per pixel, 2018–2024)
146
177
 
147
178
  | Method | Mean Slope | Std | Correlation with Raw |
148
179
  |--------|:----------:|:---:|:--------------------:|
@@ -150,13 +181,11 @@ Validated on four Sentinel-2 NDVI scenes (March–April, 2018–2020–2022–20
150
181
  | Min-Max | 0.027454 | 0.2170 | r = 0.511 |
151
182
  | **Z-Split** | **−0.000120** | **0.0167** | **r = 0.995** |
152
183
 
153
- Z-Split preserves the original temporal signal with **99.5% fidelity**. Min-Max introduces artificial trends with slope Std **13× larger** than raw data.
184
+ #### C. Multi-Temporal Change Maps
154
185
 
155
- #### B. Anomaly Detection (2024 vs. multi-year mean)
186
+ The figure below compares Linear Trend, Anomaly, and Coefficient of Variation maps across all three methods. Z-Split maps are visually and statistically consistent with raw data; Min-Max maps show spatially inverted anomaly patterns and 5.5× inflated variability.
156
187
 
157
- Min-Max anomaly maps show spatially inverted patterns relative to raw data. Z-Split anomaly maps are visually and statistically consistent with raw data.
158
-
159
- #### C. Coefficient of Variation (CV across 2018–2024)
188
+ ![Change Analysis](https://raw.githubusercontent.com/aalmoadi/endwi/main/Figures/change_analysis.png)
160
189
 
161
190
  | Method | Mean CV |
162
191
  |--------|:-------:|
@@ -164,14 +193,12 @@ Min-Max anomaly maps show spatially inverted patterns relative to raw data. Z-Sp
164
193
  | **Z-Split** | **0.173** |
165
194
  | Min-Max | 0.832 |
166
195
 
167
- Min-Max inflates temporal variability by **5.5×**. Z-Split stays within 15% of the raw baseline.
168
-
169
196
  #### D. Threshold Stability (Otsu across 11,424 patches)
170
197
 
171
198
  | Method | Threshold Std |
172
199
  |--------|:-------------:|
173
200
  | Min-Max | 0.0882 |
174
- | **Z-Split** | **0.2220** |
201
+ | Z-Split | 0.2220 |
175
202
  | Z-Score | 1.2012 |
176
203
 
177
204
  > **Note:** For strongly unipolar NDVI in arid urban areas, Min-Max threshold stability is comparable to Z-Split. Z-Split's primary advantage is in bipolar indices with narrow near-zero distributions (NDWI, InSAR LOS), consistent with its design objective.
@@ -182,7 +209,7 @@ Min-Max inflates temporal variability by **5.5×**. Z-Split stays within 15% of
182
209
 
183
210
  Most effective when:
184
211
  - Values cluster near zero (NDWI in arid regions, InSAR LOS over stable terrain)
185
- - Multi-temporal consistency is required
212
+ - Multi-temporal consistency is required (change detection, time series analysis)
186
213
  - Otsu or zero-based thresholding is applied
187
214
  - Cross-scene or cross-sensor comparisons are performed
188
215
 
@@ -199,14 +226,6 @@ Most effective when:
199
226
 
200
227
  ---
201
228
 
202
- ## Figures
203
-
204
- **Figure 1:** Otsu threshold distribution across 11,424 spatial patches — Min-Max vs Z-Score vs Z-Split.
205
-
206
- **Figure 2:** Multi-temporal NDVI change analysis (Linear Trend, Anomaly, CV) — Raw vs Min-Max vs Z-Split, Riyadh 2018–2024.
207
-
208
- ---
209
-
210
229
  ## Author
211
230
 
212
231
  **Abdulrhman Almoadi**
{zsplit-1.0.1 → zsplit-1.0.2}/README.md RENAMED
@@ -1,6 +1,6 @@
1
1
  # Z-Split Normalization (Zero-Preserving Split Normalization)
2
2
 
3
- **Version:** 1.0.0
3
+ **Version:** 1.0.1
4
4
  **License:** MIT
5
5
 
6
6
  ---
@@ -74,6 +74,24 @@ result = normalize(your_bipolar_index)
74
74
 
75
75
  ---
76
76
 
77
+ ## Practical Applications in Remote Sensing
78
+
79
+ Z-Split addresses a critical problem in any workflow involving bipolar indices across time or sensors:
80
+
81
+ **1. Change Detection**
82
+ Applying Min-Max before change detection introduces artificial trends that appear as land cover change but reflect only normalization artifacts. A report showing "15% vegetation decline" may be entirely attributable to normalization, not reality.
83
+
84
+ **2. Time Series Analysis**
85
+ Studies tracking floods, drought, or urban expansion over time require a clean temporal signal. Min-Max corrupts this signal by rescaling each scene independently to [0, 1], making year-to-year comparisons unreliable.
86
+
87
+ **3. Multi-temporal Machine Learning**
88
+ Models trained on Min-Max normalized multi-year data learn spurious patterns introduced by normalization rather than real spectral change. Z-Split ensures the training signal reflects actual surface conditions.
89
+
90
+ **4. InSAR and Displacement Monitoring**
91
+ In LOS displacement data, zero means no movement. Shifting zero through Min-Max misclassifies stable pixels as deforming, directly corrupting subsidence or uplift maps.
92
+
93
+ ---
94
+
77
95
  ## Validation
78
96
 
79
97
  ### 1. NDWI — Optical Remote Sensing
@@ -84,7 +102,20 @@ Tested on Sentinel-2 NDWI data. Z-Split successfully expanded near-zero clustere
84
102
 
85
103
  Validated on four Sentinel-2 NDVI scenes (March–April, 2018–2020–2022–2024) over Riyadh, Saudi Arabia (scale: 10 m, 5603 × 5129 pixels per scene) across 11,424 spatial patches.
86
104
 
87
- #### A. Linear Trend (Slope per pixel, 2018–2024)
105
+ #### A. Normalization Artifact Analysis
106
+
107
+ The figure below shows the deviation of each normalization method from the raw temporal signal. Min-Max introduces large artificial trends across the entire scene (colored map). Z-Split introduces virtually no deviation (near-white map).
108
+
109
+ ![Deviation Analysis](https://raw.githubusercontent.com/aalmoadi/endwi/main/Figures/deviation_analysis.png)
110
+
111
+ | Method | Mean Absolute Deviation | Std |
112
+ |--------|:-----------------------:|:---:|
113
+ | Min-Max | 0.1858 | 0.2091 |
114
+ | **Z-Split** | **0.0013** | **0.0017** |
115
+
116
+ Z-Split preserves the temporal signal with **143× greater fidelity** than Min-Max.
117
+
118
+ #### B. Linear Trend (Slope per pixel, 2018–2024)
88
119
 
89
120
  | Method | Mean Slope | Std | Correlation with Raw |
90
121
  |--------|:----------:|:---:|:--------------------:|
@@ -92,13 +123,11 @@ Validated on four Sentinel-2 NDVI scenes (March–April, 2018–2020–2022–20
92
123
  | Min-Max | 0.027454 | 0.2170 | r = 0.511 |
93
124
  | **Z-Split** | **−0.000120** | **0.0167** | **r = 0.995** |
94
125
 
95
- Z-Split preserves the original temporal signal with **99.5% fidelity**. Min-Max introduces artificial trends with slope Std **13× larger** than raw data.
126
+ #### C. Multi-Temporal Change Maps
96
127
 
97
- #### B. Anomaly Detection (2024 vs. multi-year mean)
128
+ The figure below compares Linear Trend, Anomaly, and Coefficient of Variation maps across all three methods. Z-Split maps are visually and statistically consistent with raw data; Min-Max maps show spatially inverted anomaly patterns and 5.5× inflated variability.
98
129
 
99
- Min-Max anomaly maps show spatially inverted patterns relative to raw data. Z-Split anomaly maps are visually and statistically consistent with raw data.
100
-
101
- #### C. Coefficient of Variation (CV across 2018–2024)
130
+ ![Change Analysis](https://raw.githubusercontent.com/aalmoadi/endwi/main/Figures/change_analysis.png)
102
131
 
103
132
  | Method | Mean CV |
104
133
  |--------|:-------:|
@@ -106,14 +135,12 @@ Min-Max anomaly maps show spatially inverted patterns relative to raw data. Z-Sp
106
135
  | **Z-Split** | **0.173** |
107
136
  | Min-Max | 0.832 |
108
137
 
109
- Min-Max inflates temporal variability by **5.5×**. Z-Split stays within 15% of the raw baseline.
110
-
111
138
  #### D. Threshold Stability (Otsu across 11,424 patches)
112
139
 
113
140
  | Method | Threshold Std |
114
141
  |--------|:-------------:|
115
142
  | Min-Max | 0.0882 |
116
- | **Z-Split** | **0.2220** |
143
+ | Z-Split | 0.2220 |
117
144
  | Z-Score | 1.2012 |
118
145
 
119
146
  > **Note:** For strongly unipolar NDVI in arid urban areas, Min-Max threshold stability is comparable to Z-Split. Z-Split's primary advantage is in bipolar indices with narrow near-zero distributions (NDWI, InSAR LOS), consistent with its design objective.
@@ -124,7 +151,7 @@ Min-Max inflates temporal variability by **5.5×**. Z-Split stays within 15% of
124
151
 
125
152
  Most effective when:
126
153
  - Values cluster near zero (NDWI in arid regions, InSAR LOS over stable terrain)
127
- - Multi-temporal consistency is required
154
+ - Multi-temporal consistency is required (change detection, time series analysis)
128
155
  - Otsu or zero-based thresholding is applied
129
156
  - Cross-scene or cross-sensor comparisons are performed
130
157
 
@@ -141,14 +168,6 @@ Most effective when:
141
168
 
142
169
  ---
143
170
 
144
- ## Figures
145
-
146
- **Figure 1:** Otsu threshold distribution across 11,424 spatial patches — Min-Max vs Z-Score vs Z-Split.
147
-
148
- **Figure 2:** Multi-temporal NDVI change analysis (Linear Trend, Anomaly, CV) — Raw vs Min-Max vs Z-Split, Riyadh 2018–2024.
149
-
150
- ---
151
-
152
171
  ## Author
153
172
 
154
173
  **Abdulrhman Almoadi**
{zsplit-1.0.1 → zsplit-1.0.2}/pyproject.toml RENAMED
@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
4
4
 
5
5
  [project]
6
6
  name = "zsplit"
7
- version = "1.0.1"
7
+ version = "1.0.2"
8
8
  description = "Z-Split Normalization: Zero-Preserving Split Normalization for bipolar spectral indices"
9
9
  readme = "README.md"
10
10
  license = { file = "LICENSE" }
{zsplit-1.0.1 → zsplit-1.0.2/zsplit.egg-info}/PKG-INFO RENAMED
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: zsplit
3
- Version: 1.0.1
3
+ Version: 1.0.2
4
4
  Summary: Z-Split Normalization: Zero-Preserving Split Normalization for bipolar spectral indices
5
5
  Author-email: Abdulrhman Almoadi <aalmoadi@kacst.edu.sa>
6
6
  License: MIT License
@@ -58,7 +58,7 @@ Dynamic: license-file
58
58
 
59
59
  # Z-Split Normalization (Zero-Preserving Split Normalization)
60
60
 
61
- **Version:** 1.0.0
61
+ **Version:** 1.0.1
62
62
  **License:** MIT
63
63
 
64
64
  ---
@@ -132,6 +132,24 @@ result = normalize(your_bipolar_index)
132
132
 
133
133
  ---
134
134
 
135
+ ## Practical Applications in Remote Sensing
136
+
137
+ Z-Split addresses a critical problem in any workflow involving bipolar indices across time or sensors:
138
+
139
+ **1. Change Detection**
140
+ Applying Min-Max before change detection introduces artificial trends that appear as land cover change but reflect only normalization artifacts. A report showing "15% vegetation decline" may be entirely attributable to normalization, not reality.
141
+
142
+ **2. Time Series Analysis**
143
+ Studies tracking floods, drought, or urban expansion over time require a clean temporal signal. Min-Max corrupts this signal by rescaling each scene independently to [0, 1], making year-to-year comparisons unreliable.
144
+
145
+ **3. Multi-temporal Machine Learning**
146
+ Models trained on Min-Max normalized multi-year data learn spurious patterns introduced by normalization rather than real spectral change. Z-Split ensures the training signal reflects actual surface conditions.
147
+
148
+ **4. InSAR and Displacement Monitoring**
149
+ In LOS displacement data, zero means no movement. Shifting zero through Min-Max misclassifies stable pixels as deforming, directly corrupting subsidence or uplift maps.
150
+
151
+ ---
152
+
135
153
  ## Validation
136
154
 
137
155
  ### 1. NDWI — Optical Remote Sensing
@@ -142,7 +160,20 @@ Tested on Sentinel-2 NDWI data. Z-Split successfully expanded near-zero clustere
142
160
 
143
161
  Validated on four Sentinel-2 NDVI scenes (March–April, 2018–2020–2022–2024) over Riyadh, Saudi Arabia (scale: 10 m, 5603 × 5129 pixels per scene) across 11,424 spatial patches.
144
162
 
145
- #### A. Linear Trend (Slope per pixel, 2018–2024)
163
+ #### A. Normalization Artifact Analysis
164
+
165
+ The figure below shows the deviation of each normalization method from the raw temporal signal. Min-Max introduces large artificial trends across the entire scene (colored map). Z-Split introduces virtually no deviation (near-white map).
166
+
167
+ ![Deviation Analysis](https://raw.githubusercontent.com/aalmoadi/endwi/main/Figures/deviation_analysis.png)
168
+
169
+ | Method | Mean Absolute Deviation | Std |
170
+ |--------|:-----------------------:|:---:|
171
+ | Min-Max | 0.1858 | 0.2091 |
172
+ | **Z-Split** | **0.0013** | **0.0017** |
173
+
174
+ Z-Split preserves the temporal signal with **143× greater fidelity** than Min-Max.
175
+
176
+ #### B. Linear Trend (Slope per pixel, 2018–2024)
146
177
 
147
178
  | Method | Mean Slope | Std | Correlation with Raw |
148
179
  |--------|:----------:|:---:|:--------------------:|
@@ -150,13 +181,11 @@ Validated on four Sentinel-2 NDVI scenes (March–April, 2018–2020–2022–20
150
181
  | Min-Max | 0.027454 | 0.2170 | r = 0.511 |
151
182
  | **Z-Split** | **−0.000120** | **0.0167** | **r = 0.995** |
152
183
 
153
- Z-Split preserves the original temporal signal with **99.5% fidelity**. Min-Max introduces artificial trends with slope Std **13× larger** than raw data.
184
+ #### C. Multi-Temporal Change Maps
154
185
 
155
- #### B. Anomaly Detection (2024 vs. multi-year mean)
186
+ The figure below compares Linear Trend, Anomaly, and Coefficient of Variation maps across all three methods. Z-Split maps are visually and statistically consistent with raw data; Min-Max maps show spatially inverted anomaly patterns and 5.5× inflated variability.
156
187
 
157
- Min-Max anomaly maps show spatially inverted patterns relative to raw data. Z-Split anomaly maps are visually and statistically consistent with raw data.
158
-
159
- #### C. Coefficient of Variation (CV across 2018–2024)
188
+ ![Change Analysis](https://raw.githubusercontent.com/aalmoadi/endwi/main/Figures/change_analysis.png)
160
189
 
161
190
  | Method | Mean CV |
162
191
  |--------|:-------:|
@@ -164,14 +193,12 @@ Min-Max anomaly maps show spatially inverted patterns relative to raw data. Z-Sp
164
193
  | **Z-Split** | **0.173** |
165
194
  | Min-Max | 0.832 |
166
195
 
167
- Min-Max inflates temporal variability by **5.5×**. Z-Split stays within 15% of the raw baseline.
168
-
169
196
  #### D. Threshold Stability (Otsu across 11,424 patches)
170
197
 
171
198
  | Method | Threshold Std |
172
199
  |--------|:-------------:|
173
200
  | Min-Max | 0.0882 |
174
- | **Z-Split** | **0.2220** |
201
+ | Z-Split | 0.2220 |
175
202
  | Z-Score | 1.2012 |
176
203
 
177
204
  > **Note:** For strongly unipolar NDVI in arid urban areas, Min-Max threshold stability is comparable to Z-Split. Z-Split's primary advantage is in bipolar indices with narrow near-zero distributions (NDWI, InSAR LOS), consistent with its design objective.
@@ -182,7 +209,7 @@ Min-Max inflates temporal variability by **5.5×**. Z-Split stays within 15% of
182
209
 
183
210
  Most effective when:
184
211
  - Values cluster near zero (NDWI in arid regions, InSAR LOS over stable terrain)
185
- - Multi-temporal consistency is required
212
+ - Multi-temporal consistency is required (change detection, time series analysis)
186
213
  - Otsu or zero-based thresholding is applied
187
214
  - Cross-scene or cross-sensor comparisons are performed
188
215
 
@@ -199,14 +226,6 @@ Most effective when:
199
226
 
200
227
  ---
201
228
 
202
- ## Figures
203
-
204
- **Figure 1:** Otsu threshold distribution across 11,424 spatial patches — Min-Max vs Z-Score vs Z-Split.
205
-
206
- **Figure 2:** Multi-temporal NDVI change analysis (Linear Trend, Anomaly, CV) — Raw vs Min-Max vs Z-Split, Riyadh 2018–2024.
207
-
208
- ---
209
-
210
229
  ## Author
211
230
 
212
231
  **Abdulrhman Almoadi**
{zsplit-1.0.1 → zsplit-1.0.2}/LICENSE RENAMED
File without changes
{zsplit-1.0.1 → zsplit-1.0.2}/setup.cfg RENAMED
File without changes
{zsplit-1.0.1 → zsplit-1.0.2}/zsplit/__init__.py RENAMED
File without changes
{zsplit-1.0.1 → zsplit-1.0.2}/zsplit/core.py RENAMED
File without changes