Upload readme.md
Browse files- config/readme.md +165 -0
config/readme.md
ADDED
@@ -0,0 +1,165 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Configuration File Instructions
|
2 |
+
|
3 |
+
Below are the instructions to create a valid configuration file.
|
4 |
+
|
5 |
+
## Configuration File Structure
|
6 |
+
|
7 |
+
The configuration file should be in YAML format (.config file) and follow the structure outlined below for **ICU Data**:
|
8 |
+
|
9 |
+
```yaml
|
10 |
+
disease_label: <disease_label> mandatory only if the prediction task is Phenotype
|
11 |
+
timePrediction: <timePrediction>
|
12 |
+
timeWindow: <timeWindow>
|
13 |
+
timebucket: <timebucket>
|
14 |
+
radimp: <radimp>
|
15 |
+
predW: <predW>
|
16 |
+
diagnosis: <diagnosis>
|
17 |
+
output: <output>
|
18 |
+
chart: <chart>
|
19 |
+
proc: <proc>
|
20 |
+
meds: <meds>
|
21 |
+
disease_filter: <disease_filter>
|
22 |
+
icu_no_icu: <icu_no_icu>
|
23 |
+
groupingDiag: <groupingDiag>
|
24 |
+
select_diag: <select_diag>
|
25 |
+
select_med: <select_med>
|
26 |
+
select_proc: <select_proc>
|
27 |
+
select_out: <select_out>
|
28 |
+
select_chart: <select_chart>
|
29 |
+
outlier_removal: <outlier_removal>
|
30 |
+
outlier: <outlier>
|
31 |
+
left_outlier: <left_outlier>
|
32 |
+
```
|
33 |
+
|
34 |
+
The configuration file should be in YAML format (.config file) and follow the structure outlined below for **Non-ICU Data**:
|
35 |
+
|
36 |
+
```yaml
|
37 |
+
disease_label: <disease_label> mandatory only if the prediction task is Phenotype
|
38 |
+
timePrediction: <timePrediction>
|
39 |
+
timeWindow: <timeWindow>
|
40 |
+
timebucket: <timebucket>
|
41 |
+
radimp: <radimp>
|
42 |
+
predW: <predW>
|
43 |
+
diagnosis: <diagnosis>
|
44 |
+
lab: <lab>
|
45 |
+
proc: <proc>
|
46 |
+
meds: <meds>
|
47 |
+
disease_filter: <disease_filter>
|
48 |
+
icu_no_icu: <icu_no_icu>
|
49 |
+
groupingDiag: <groupingDiag>
|
50 |
+
groupingProc: <groupingProc>
|
51 |
+
groupingMed: <groupingMed>
|
52 |
+
select_diag: <select_diag>
|
53 |
+
select_med: <select_med>
|
54 |
+
select_proc: <select_proc>
|
55 |
+
select_lab: <select_lab>
|
56 |
+
outlier_removal: <outlier_removal>
|
57 |
+
outlier: <outlier>
|
58 |
+
left_outlier: <left_outlier>
|
59 |
+
```
|
60 |
+
|
61 |
+
Replace the `<variable>` placeholders with the corresponding values specific to your use case. Detailed explanations of each variable and their valid values are provided in the next section.
|
62 |
+
|
63 |
+
## Variable Definitions and Valid Values
|
64 |
+
|
65 |
+
- `disease_label` (string): Specifies the disease label for Phenotype prediction task. Don't provide the line if the task is not Phenotype. Valid values: CAD, Heart Failure, CKD, COPD.
|
66 |
+
- `timePrediction` (integer): Specifies the time prediction (days). Valid values depend on the task:
|
67 |
+
- For Phenotype task: 30
|
68 |
+
- For Mortality task: 0
|
69 |
+
- For Length of Stay task: Between 1 and 10 (inclusive)
|
70 |
+
- For Readmission task: Between 10 and 150 (inclusive), multiple of 10
|
71 |
+
- `timeWindow` (string): Specifies the time window. Valid values:
|
72 |
+
- For Phenotype or Readmission task: Last X hours (with 24 <= X >= 72)
|
73 |
+
- For Mortality or Length of Stay tasks: First X hours (with 24 <= X >= 72)
|
74 |
+
- `timebucket` (integer): Specifies the time bucket. Valid values: Between 1 and 6 (inclusive).
|
75 |
+
- `radimp` (string): Specifies the imputation method. Valid values:
|
76 |
+
- No Imputation
|
77 |
+
- forward fill and mean
|
78 |
+
- forward fill and median
|
79 |
+
- `predW` (integer): Specifies the prediction window. Valid values depend on the task:
|
80 |
+
- For Phenotype, Length of Stay, or Readmission tasks: 0
|
81 |
+
- For Mortality task: Between 2 and 8 (inclusive)
|
82 |
+
- `diagnosis`, `output`, `chart`, `proc`, `meds`, `lab` (boolean): Specifies whether to include each respective feature. Valid values: True or False.
|
83 |
+
- `disease_filter` (string): Specifies the disease filter if focusing on a cohort with a specific chronic disease. Valid values:
|
84 |
+
- Heart Failure
|
85 |
+
- COPD
|
86 |
+
- CKD
|
87 |
+
- CAD
|
88 |
+
- No Disease Filter
|
89 |
+
- `icu_no_icu` (string): Specifies the dataset type. Valid values: ICU.
|
90 |
+
- `groupingDiag` (string): Specifies the grouping ICD option for diagnosis. Valid values:
|
91 |
+
- Convert ICD-9 to ICD-10 and group ICD-10 codes
|
92 |
+
- Keep both ICD-9 and ICD-10 codes
|
93 |
+
- Convert ICD-9 to ICD-10 codes
|
94 |
+
- `groupingMed`: Specifies if grouping Medication codes should be done to use Non propietary names : Valid values : Yes, No
|
95 |
+
- `groupingProc` : Specifies the ICD codes version to perform grouping for procedures : Valid values :
|
96 |
+
- ICD-9 and ICD-10
|
97 |
+
- ICD-10
|
98 |
+
- `select_diag`, `select_med`, `select_proc`, `select_out`, `select_chart`,`select_lab` (boolean): Specifies whether to do features selection as describe in https://github.com/healthylaife/MIMIC-IV-Data-Pipeline benchmark. Valid values: True or False.
|
99 |
+
- `outlier_removal` (string): Specifies the outlier removal method. Valid values:
|
100 |
+
- No outlier detection
|
101 |
+
- Impute Outlier (default:98)
|
102 |
+
- Remove outliers (default:98)
|
103 |
+
- `outlier` (integer): Specifies the outlier threshold. Valid values: Between 90 and 99 (inclusive).
|
104 |
+
- `left_outlier` (integer): Specifies the left outlier threshold. Valid values: Between 0 and 10 (inclusive).
|
105 |
+
|
106 |
+
## Example Configuration File
|
107 |
+
|
108 |
+
Here's an example of a valid configuration file for **ICU Data**:
|
109 |
+
|
110 |
+
```yaml
|
111 |
+
disease_label: CAD
|
112 |
+
timePrediction: 30
|
113 |
+
timeWindow: Last 72 hours
|
114 |
+
timebucket: 2
|
115 |
+
radimp: forward fill and mean
|
116 |
+
predW: 0
|
117 |
+
diagnosis: True
|
118 |
+
output: True
|
119 |
+
chart: True
|
120 |
+
proc: True
|
121 |
+
meds: True
|
122 |
+
disease_filter: No Disease Filter
|
123 |
+
icu_no_icu: ICU
|
124 |
+
groupingICD: Convert ICD-9 to ICD-10 and group ICD-10 codes
|
125 |
+
select_diag: False
|
126 |
+
select_med: False
|
127 |
+
select_proc: False
|
128 |
+
select_out: False
|
129 |
+
select_chart: False
|
130 |
+
outlier_removal: Impute Outlier (default:98)
|
131 |
+
outlier: 98
|
132 |
+
left_outlier: 0
|
133 |
+
```
|
134 |
+
Here's an example of a valid configuration file for **Non-ICU Data**:
|
135 |
+
|
136 |
+
```yaml
|
137 |
+
timePrediction: 0
|
138 |
+
timeWindow: First 48 hours
|
139 |
+
timebucket: 2
|
140 |
+
radimp: forward fill and mean
|
141 |
+
predW: 2
|
142 |
+
diagnosis: True
|
143 |
+
lab: True
|
144 |
+
proc: False
|
145 |
+
meds: False
|
146 |
+
disease_filter: CKD
|
147 |
+
icu_no_icu: ICU
|
148 |
+
groupingDiag: Convert ICD-9 to ICD-10 and group ICD-10 codes
|
149 |
+
select_diag: False
|
150 |
+
select_med: False
|
151 |
+
select_proc: False
|
152 |
+
select_lab: False
|
153 |
+
outlier_removal: Impute Outlier (default:98)
|
154 |
+
outlier: 98
|
155 |
+
left_outlier: 0
|
156 |
+
groupingMed: Yes
|
157 |
+
groupingProc: ICD-10
|
158 |
+
```
|
159 |
+
|
160 |
+
Feel free to modify the values to fit your specific requirements.
|
161 |
+
|
162 |
+
## Usage
|
163 |
+
Provide the full path of your configuration file while calling the loading dataset function with the parameter config_path=`<path_to_config_file>`.
|
164 |
+
|
165 |
+
For more understanding of the configuration please refer to https://github.com/healthylaife/MIMIC-IV-Data-Pipeline.
|