bdext 0.1.64__py3-none-any.whl → 0.1.66__py3-none-any.whl

README.md

@@ -1,149 +1,211 @@
-# bdeissct_dl
+# bdext
 
-Estimator of BDEISS-CT model parameters from phylogenetic trees 
+The bdext package provides scripts to train and assess
+Deep-Learning-enables estimators of BD(EI)(SS)(CT) model parameters from phylogenetic trees 
 
 
 
 [//]: # ([![DOI:10.1093/sysbio/syad059](https://zenodo.org/badge/DOI/10.1093/sysbio/syad059.svg)](https://doi.org/10.1093/sysbio/syad059))
-[![GitHub release](https://img.shields.io/github/v/release/evolbioinfo/bdeissct_dl.svg)](https://github.com/evolbioinfo/bdeissct_dl/releases)
-[![PyPI version](https://badge.fury.io/py/bdeissct_dl.svg)](https://pypi.org/project/bdeissct_dl/)
-[![PyPI downloads](https://shields.io/pypi/dm/bdeissct_dl)](https://pypi.org/project/bdeissct_dl/)
-[![Docker pulls](https://img.shields.io/docker/pulls/evolbioinfo/bdeissct)](https://hub.docker.com/r/evolbioinfo/bdeissct/tags)
+[//]: # ([![GitHub release](https://img.shields.io/github/v/release/evolbioinfo/bdext.svg)](https://github.com/evolbioinfo/bdext/releases))
+[![PyPI version](https://badge.fury.io/py/bdext.svg)](https://pypi.org/project/bdext/)
+[![PyPI downloads](https://shields.io/pypi/dm/bdext)](https://pypi.org/project/bdext)
+[![Docker pulls](https://img.shields.io/docker/pulls/evolbioinfo/bdext)](https://hub.docker.com/r/evolbioinfo/bdext/tags)
 
 ## BDEISS-CT model
 
-BD-PN model extends the classical birth-death (BD) model with incomplete sampling [[Stadler 2009]](https://pubmed.ncbi.nlm.nih.gov/19631666/), by adding partner notification (PN).
-Under this model, infected individuals can transmit their pathogen with a constant rate λ, 
-get removed (become non-infectious) with a constant rate ψ, 
-and their pathogen can be sampled upon removal 
-with a constant probability ρ. On top of that, in the BD-PN model, 
-at the moment of sampling the sampled individual 
-might notify their most recent partner with a constant probability υ. 
-Upon notification, the partner is removed almost instantaneously (modeled via a constant notified
-removal rate φ >> ψ) and their pathogen is sampled.
+The Birth-Death (BD) Exposed-Infectious (EI) with SuperSpreading (SS) and Contact-Tracing (CT) model (BDEISS-CT) 
+can be described with the following 8 parameters:
 
-BD-PN model therefore has 5 parameters:
-* λ -- transmission rate
-* ψ -- removal rate
-* ρ -- sampling probability upon removal
-* υ -- probability to notify the last partner upon sampling
-* φ -- removal (and sampling) rate after notification
+* average reproduction number R;
+* average total infection duration d;
+* incubation period d<sub>inc</sub>;
+* sampling probability ρ;
+* fraction of superspreaders f<sub>S</sub>;
+* super-spreading transmission increase X<sub>S</sub>;
+* contact tracing probability υ;
+* contact-traced removal speed up X<sub>C</sub>.
 
-These parameters can be expressed in terms of the following epidemiological parameters:
-* R<sub>0</sub>=λ/ψ -- reproduction number
-* 1/ψ -- infectious time
-* 1/φ -- partner removal time
+Setting d<sub>inc</sub>=0 removes incubation (EI), setting f<sub>S</sub>=0 removes superspreading (SS), while setting υ=0 removes contact-tracing (CT).
 
-BD-CT model makes 3 assumptions:
-1. only observed individuals can notify (instead of any removed individual);
-2. notified individuals are always observed upon removal;
-3. only the most recent partner can get notified.
+For identifiability, we require the sampling probability ρ to be given by the user. 
+The other parameters are estimated from a time-scaled phylogenetic tree.
 
-For identifiability, BD-PN model requires one of the three BD model parameters (λ, ψ, ρ) to be fixed.
+[//]: # (## BDEISS-CT parameter estimator)
 
-## BDEISS-CT parameter estimator
+[//]: # ()
+[//]: # (The bdeissct_dl package provides deep-learning-based BDEISS-CT model parameter estimator )
 
-The bdeissct_dl package provides deep-learning-based BDEISS-CT model parameter estimator 
-from a user-supplied time-scaled phylogenetic tree. 
-User must also provide a value for one of the three BD model parameters (λ, ψ, or ρ). 
-We recommend providing the sampling probability ρ, 
-which could be estimated as the number of tree tips divided by the number of declared cases for the same time period.
+[//]: # (from a user-supplied time-scaled phylogenetic tree. )
 
+[//]: # (User must also provide a value for one of the three BD model parameters &#40;λ, ψ, or ρ&#41;. )
 
-## Input data
-One needs to supply a time-scaled phylogenetic tree in newick format. 
-In the examples below we will use an HIV tree reconstructed from 200 sequences, 
-published in [[Rasmussen _et al._ PLoS Comput. Biol. 2017]](https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1005448), 
-which you can find at [PairTree GitHub](https://github.com/davidrasm/PairTree) 
-and in [hiv_zurich/Zurich.nwk](hiv_zurich/Zurich.nwk). 
+[//]: # (We recommend providing the sampling probability ρ, )
 
-## Installation
+[//]: # (which could be estimated as the number of tree tips divided by the number of declared cases for the same time period.)
 
-There are 4 alternative ways to run __bdeissct_dl__ on your computer: 
-with [docker](https://www.docker.com/community-edition), 
-[apptainer](https://apptainer.org/),
-in Python3, or via command line (requires installation with Python3).
+[//]: # ()
+[//]: # ()
+[//]: # (## Input data)
 
+[//]: # (One needs to supply a time-scaled phylogenetic tree in newick format. )
 
+[//]: # (In the examples below we will use an HIV tree reconstructed from 200 sequences, )
 
-### Run in python3 or command-line (for linux systems, recommended Ubuntu 21 or newer versions)
+[//]: # (published in [[Rasmussen _et al._ PLoS Comput. Biol. 2017]]&#40;https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1005448&#41;, )
 
-You could either install python (version 3.9 or higher) system-wide and then install bdeissct_dl via pip:
-```bash
-sudo apt install -y python3 python3-pip python3-setuptools python3-distutils
-pip3 install bdeissct_dl
-```
+[//]: # (which you can find at [PairTree GitHub]&#40;https://github.com/davidrasm/PairTree&#41; )
 
-or alternatively, you could install python (version 3.9 or higher) and bdeissct_dl via [conda](https://conda.io/docs/) (make sure that conda is installed first). 
-Here we will create a conda environment called _phyloenv_:
-```bash
-conda create --name phyloenv python=3.12
-conda activate phyloenv
-pip install bdeissct_dl
-```
+[//]: # (and in [hiv_zurich/Zurich.nwk]&#40;hiv_zurich/Zurich.nwk&#41;. )
 
+[//]: # ()
+[//]: # (## Installation)
 
-#### Basic usage in a command line
-If you installed __bdeissct_dl__ in a conda environment (here named _phyloenv_), do not forget to first activate it, e.g.
+[//]: # ()
+[//]: # (There are 4 alternative ways to run __bdeissct_dl__ on your computer: )
 
-```bash
-conda activate phyloenv
-```
+[//]: # (with [docker]&#40;https://www.docker.com/community-edition&#41;, )
 
-Run the following command to estimate the BDEISS_CT parameters and their 95% CIs for this tree, assuming the sampling probability of 0.25, 
-and save the estimated parameters to a comma-separated file estimates.csv.
-```bash
-bdeissct_infer --nwk Zurich.nwk --ci --p 0.25 --log estimates.csv
-```
+[//]: # ([apptainer]&#40;https://apptainer.org/&#41;,)
 
-#### Help
+[//]: # (in Python3, or via command line &#40;requires installation with Python3&#41;.)
 
-To see detailed options, run:
-```bash
-bdeissct_infer --help
-```
+[//]: # ()
+[//]: # ()
+[//]: # ()
+[//]: # (### Run in python3 or command-line &#40;for linux systems, recommended Ubuntu 21 or newer versions&#41;)
 
+[//]: # ()
+[//]: # (You could either install python &#40;version 3.9 or higher&#41; system-wide and then install bdeissct_dl via pip:)
 
-### Run with docker
+[//]: # (```bash)
 
-#### Basic usage
-Once [docker](https://www.docker.com/community-edition) is installed, 
-run the following command to estimate BDEISS-CT model parameters:
-```bash
-docker run -v <path_to_the_folder_containing_the_tree>:/data:rw -t evolbioinfo/bdeissct --nwk /data/Zurich.nwk --ci --p 0.25 --log /data/estimates.csv
-```
+[//]: # (sudo apt install -y python3 python3-pip python3-setuptools python3-distutils)
 
-This will produce a comma-separated file estimates.csv in the <path_to_the_folder_containing_the_tree> folder,
- containing the estimated parameter values and their 95% CIs (can be viewed with a text editor, Excel or Libre Office Calc).
+[//]: # (pip3 install bdeissct_dl)
 
-#### Help
+[//]: # (```)
 
-To see advanced options, run
-```bash
-docker run -t evolbioinfo/bdeissct -h
-```
+[//]: # ()
+[//]: # (or alternatively, you could install python &#40;version 3.9 or higher&#41; and bdeissct_dl via [conda]&#40;https://conda.io/docs/&#41; &#40;make sure that conda is installed first&#41;. )
 
+[//]: # (Here we will create a conda environment called _phyloenv_:)
 
+[//]: # (```bash)
 
-### Run with apptainer
+[//]: # (conda create --name phyloenv python=3.12)
 
-#### Basic usage
-Once [apptainer](https://apptainer.org/docs/user/latest/quick_start.html#installation) is installed, 
-run the following command to estimate BDEISS-CT model parameters (from the folder where the Zurich.nwk tree is contained):
+[//]: # (conda activate phyloenv)
 
-```bash
-apptainer run docker://evolbioinfo/bdeissct --nwk Zurich.nwk --ci --p 0.25 --log estimates.csv
-```
+[//]: # (pip install bdeissct_dl)
 
-This will produce a comma-separated file estimates.csv,
- containing the estimated parameter values and their 95% CIs (can be viewed with a text editor, Excel or Libre Office Calc).
+[//]: # (```)
 
+[//]: # ()
+[//]: # ()
+[//]: # (#### Basic usage in a command line)
 
-#### Help
+[//]: # (If you installed __bdeissct_dl__ in a conda environment &#40;here named _phyloenv_&#41;, do not forget to first activate it, e.g.)
 
-To see advanced options, run
-```bash
-apptainer run docker://evolbioinfo/bdeissct -h
-```
+[//]: # ()
+[//]: # (```bash)
 
+[//]: # (conda activate phyloenv)
 
+[//]: # (```)
+
+[//]: # ()
+[//]: # (Run the following command to estimate the BDEISS_CT parameters and their 95% CIs for this tree, assuming the sampling probability of 0.25, )
+
+[//]: # (and save the estimated parameters to a comma-separated file estimates.csv.)
+
+[//]: # (```bash)
+
+[//]: # (bdeissct_infer --nwk Zurich.nwk --ci --p 0.25 --log estimates.csv)
+
+[//]: # (```)
+
+[//]: # ()
+[//]: # (#### Help)
+
+[//]: # ()
+[//]: # (To see detailed options, run:)
+
+[//]: # (```bash)
+
+[//]: # (bdeissct_infer --help)
+
+[//]: # (```)
+
+[//]: # ()
+[//]: # ()
+[//]: # (### Run with docker)
+
+[//]: # ()
+[//]: # (#### Basic usage)
+
+[//]: # (Once [docker]&#40;https://www.docker.com/community-edition&#41; is installed, )
+
+[//]: # (run the following command to estimate BDEISS-CT model parameters:)
+
+[//]: # (```bash)
+
+[//]: # (docker run -v <path_to_the_folder_containing_the_tree>:/data:rw -t evolbioinfo/bdeissct --nwk /data/Zurich.nwk --ci --p 0.25 --log /data/estimates.csv)
+
+[//]: # (```)
+
+[//]: # ()
+[//]: # (This will produce a comma-separated file estimates.csv in the <path_to_the_folder_containing_the_tree> folder,)
+
+[//]: # ( containing the estimated parameter values and their 95% CIs &#40;can be viewed with a text editor, Excel or Libre Office Calc&#41;.)
+
+[//]: # ()
+[//]: # (#### Help)
+
+[//]: # ()
+[//]: # (To see advanced options, run)
+
+[//]: # (```bash)
+
+[//]: # (docker run -t evolbioinfo/bdeissct -h)
+
+[//]: # (```)
+
+[//]: # ()
+[//]: # ()
+[//]: # ()
+[//]: # (### Run with apptainer)
+
+[//]: # ()
+[//]: # (#### Basic usage)
+
+[//]: # (Once [apptainer]&#40;https://apptainer.org/docs/user/latest/quick_start.html#installation&#41; is installed, )
+
+[//]: # (run the following command to estimate BDEISS-CT model parameters &#40;from the folder where the Zurich.nwk tree is contained&#41;:)
+
+[//]: # ()
+[//]: # (```bash)
+
+[//]: # (apptainer run docker://evolbioinfo/bdeissct --nwk Zurich.nwk --ci --p 0.25 --log estimates.csv)
+
+[//]: # (```)
+
+[//]: # ()
+[//]: # (This will produce a comma-separated file estimates.csv,)
+
+[//]: # ( containing the estimated parameter values and their 95% CIs &#40;can be viewed with a text editor, Excel or Libre Office Calc&#41;.)
+
+[//]: # ()
+[//]: # ()
+[//]: # (#### Help)
+
+[//]: # ()
+[//]: # (To see advanced options, run)
+
+[//]: # (```bash)
+
+[//]: # (apptainer run docker://evolbioinfo/bdeissct -h)
+
+[//]: # (```)
+
+[//]: # ()
+[//]: # ()

bdeissct_dl/__init__.py

@@ -7,12 +7,10 @@ warnings.filterwarnings('ignore', r'divide by zero encountered in log')
 
 
 MODEL_PATH = os.path.join(os.path.dirname(__file__), 'models')
-TRAINING_PATH = os.path.join(os.path.dirname(__file__), 'data')
-
 
 
 EPOCHS = 1000
-BATCH_SIZE = 1024
+BATCH_SIZE = 8192
 
 
 

bdeissct_dl/bdeissct_model.py

@@ -1,42 +1,26 @@
 from collections import defaultdict
 
 LA = 'la'
-LA_AVG = 'avg_la'
 PSI = 'psi'
 RHO = 'rho'
 INFECTIOUS_TIME = 'd_I'
 REPRODUCTIVE_NUMBER = 'R'
 INFECTION_DURATION = 'd'
 
-F_E = 'f_E'
 MU = 'mu'
 INCUBATION_PERIOD = 'd_E'
 
 F_S = 'f_S'
 X_S = 'X_S'
 
-F_S_X_S = 'f_S_X_S'
-
-PI_I = 'pi_I'
-PI_IC = 'pi_IC'
-PI_S = 'pi_S'
-PI_SC = 'pi_SC'
-PI_E = 'pi_E'
-PI_EC = 'pi_EC'
-
-PIS = 'pi'
-
 
 X_C = 'X_C'
 UPSILON = 'upsilon'
 
-UPS_X_C = 'upsilon_X_C'
-
 KAPPA = 'kappa'
-PHI = 'phi'
 REMOVAL_TIME_AFTER_NOTIFICATION = 'd_C'
 
-RATE_PARAMETERS = (LA, PSI, PHI, MU, LA_AVG)
+RATE_PARAMETERS = (LA, PSI, MU)
 TIME_PARAMETERS = (INCUBATION_PERIOD, INFECTIOUS_TIME, REMOVAL_TIME_AFTER_NOTIFICATION, INFECTION_DURATION)
 
 
@@ -49,27 +33,15 @@ DEFAULT_MAX_RATE = 1e3
 
 BD = 'BD'
 BDCT = 'BDCT'
-BDCT1 = 'BDCT1'
-BDCT2 = 'BDCT2'
-BDCT2000 = 'BDCT2000'
 
 BDEI = 'BDEI'
 BDEICT = 'BDEICT'
-BDEICT1 = 'BDEICT1'
-BDEICT2 = 'BDEICT2'
-BDEICT2000 = 'BDEICT2000'
 
 BDSS = 'BDSS'
 BDSSCT = 'BDSSCT'
-BDSSCT1 = 'BDSSCT1'
-BDSSCT2 = 'BDSSCT2'
-BDSSCT2000 = 'BDSSCT2000'
 
 BDEISS = 'BDEISS'
 BDEISSCT = 'BDEISSCT'
-BDEISSCT1 = 'BDEISSCT1'
-BDEISSCT2 = 'BDEISSCT2'
-BDEISSCT2000 = 'BDEISSCT2000'
 
 
 
@@ -80,16 +52,9 @@ MODELS = (BD, BDCT, \
           BDSS, BDSSCT, \
           BDEISS, BDEISSCT)
 
-DATA_TYPES = ['tree']
-
-
-WITH_PIS = False
 
-# TARGET_COLUMNS_BDCT = (REPRODUCTIVE_NUMBER, INFECTIOUS_TIME, REMOVAL_TIME_AFTER_NOTIFICATION, UPSILON)
-# TARGET_COLUMNS_BD = (REPRODUCTIVE_NUMBER, INFECTIOUS_TIME)
-# TARGET_CT_COLUMNS = (UPSILON, X_C, KAPPA)
 TARGET_CT_COLUMNS = (UPSILON, X_C)
-TARGET_INCUBATION_COLUMNS = (F_E,)
+TARGET_INCUBATION_COLUMNS = (INCUBATION_PERIOD,)
 TARGET_SS_COLUMNS = (F_S, X_S)
 TARGET_COLUMNS_BD = (REPRODUCTIVE_NUMBER, INFECTION_DURATION)
 TARGET_COLUMNS_BDCT = TARGET_COLUMNS_BD + TARGET_CT_COLUMNS
@@ -100,33 +65,14 @@ TARGET_COLUMNS_BDSSCT = TARGET_COLUMNS_BDSS + TARGET_CT_COLUMNS
 TARGET_COLUMNS_BDEISS = TARGET_COLUMNS_BDEI + TARGET_SS_COLUMNS
 TARGET_COLUMNS_BDEISSCT = TARGET_COLUMNS_BDEISS + TARGET_CT_COLUMNS
 
-if WITH_PIS:
-    TARGET_COLUMNS_BD = TARGET_COLUMNS_BD
-    TARGET_COLUMNS_BDCT = TARGET_COLUMNS_BDCT + (PI_I, PI_IC, )
-    TARGET_COLUMNS_BDEI = TARGET_COLUMNS_BDEI + (PI_E, PI_I, )
-    TARGET_COLUMNS_BDSS = TARGET_COLUMNS_BDSS + (PI_I, PI_S, )
-    TARGET_COLUMNS_BDEISS = TARGET_COLUMNS_BDEISS + (PI_E, PI_I, PI_S)
-    TARGET_COLUMNS_BDEICT = TARGET_COLUMNS_BDEICT + (PI_E, PI_I, PI_EC, PI_IC)
-    TARGET_COLUMNS_BDSSCT = TARGET_COLUMNS_BDSSCT + (PI_I, PI_S, PI_IC, PI_SC)
-    TARGET_COLUMNS_BDEISSCT = TARGET_COLUMNS_BDEISSCT + (PI_E, PI_I, PI_S, PI_EC, PI_IC, PI_SC)
-
 
 MODEL2TARGET_COLUMNS = defaultdict(lambda: TARGET_COLUMNS_BDEISSCT)
-
-MODEL2TARGET_COLUMNS[BD] = TARGET_COLUMNS_BD
-MODEL2TARGET_COLUMNS[BDEI] = TARGET_COLUMNS_BDEI
-MODEL2TARGET_COLUMNS[BDSS] = TARGET_COLUMNS_BDSS
-MODEL2TARGET_COLUMNS[BDEISS] = TARGET_COLUMNS_BDEISS
-for model in (BDCT, BDCT1, BDCT2, BDCT2000):
-    MODEL2TARGET_COLUMNS[model] = TARGET_COLUMNS_BDCT
-for model in (BDEICT, BDEICT1, BDEICT2, BDEICT2000):
-    MODEL2TARGET_COLUMNS[model] = TARGET_COLUMNS_BDEICT
-for model in (BDSSCT, BDSSCT1, BDSSCT2, BDSSCT2000):
-    MODEL2TARGET_COLUMNS[model] = TARGET_COLUMNS_BDSSCT
-for model in (BDEISSCT, BDEISSCT1, BDEISSCT2, BDEISSCT2000):
-    MODEL2TARGET_COLUMNS[model] = TARGET_COLUMNS_BDEISSCT
-
-
-
-CT_EPI_COLUMNS = [REPRODUCTIVE_NUMBER, LA, F_E, F_S, X_S, UPSILON, X_C]
-CT_RATE_COLUMNS = [PSI]
+MODEL2TARGET_COLUMNS.update({BD: TARGET_COLUMNS_BD,
+                             BDEI: TARGET_COLUMNS_BDEI,
+                             BDSS: TARGET_COLUMNS_BDSS,
+                             BDEISS: TARGET_COLUMNS_BDEISS,
+                             BDCT: TARGET_COLUMNS_BDCT,
+                             BDEICT: TARGET_COLUMNS_BDEICT,
+                             BDSSCT: TARGET_COLUMNS_BDSSCT,
+                             BDEISSCT: TARGET_COLUMNS_BDEISSCT
+                             })

bdeissct_dl/dl_model.py

@@ -1,124 +1,21 @@
 import tensorflow as tf
 from tensorflow.python.keras.utils.generic_utils import register_keras_serializable
 
-from bdeissct_dl.bdeissct_model import F_E, F_S, UPSILON, F_S_X_S, UPS_X_C, REPRODUCTIVE_NUMBER, \
-    INFECTION_DURATION, X_S, X_C, LA, PSI, RHO
+from bdeissct_dl.bdeissct_model import F_S, UPSILON, REPRODUCTIVE_NUMBER, \
+    INFECTION_DURATION, X_S, X_C, RHO, INCUBATION_PERIOD
 
 LEARNING_RATE = 0.001
 
-DELTA = 0.001
 LOSS_WEIGHTS = {
-    LA: 1,
-    PSI: 1,
     REPRODUCTIVE_NUMBER: 1,
     INFECTION_DURATION: 1,
-    UPS_X_C: 200,  # as there are 2 outputs, we multiply by 200 to scale it to [0, 200]
+    INCUBATION_PERIOD: 1,
     F_S: 200, # as it is a value between 0 and 0.5, we multiply by 200 to scale it to [0, 100]
-    F_E: 100,
     UPSILON: 100,
-    F_S_X_S: 200,  # as there are 2 outputs, we multiply by 200 to scale it to [0, 200]
     X_C: 1,
-    X_S: 1,
-    RHO: 1,
+    X_S: 1
 }
 
-QUANTILES = (0.5, )
-
-
-@register_keras_serializable(package='bdeissct_dl', name='SSLayer')
-class SSLayer(tf.keras.layers.Layer):
-    def call(self, inputs):
-        # inputs shape: (batch, 2)
-        f_S = half_sigmoid(inputs[:, 0:1])  # keepdims -> (batch, 1)
-        X_S = relu_plus_one(inputs[:, 1:2])  # (batch, 1)
-        return tf.concat([f_S, X_S], axis=-1)  # (batch, 2)
-
-    def compute_output_shape(self, input_shape):
-        # input_shape is (batch, 2) -> output_shape is (batch, 2)
-        return input_shape[:-1] + (2,)
-
-    def get_config(self):
-        # If there are no special args, only return super() config
-        return super().get_config()
-
-    @classmethod
-    def from_config(cls, config):
-        return cls(**config)
-
-
-@register_keras_serializable(package='bdeissct_dl', name='CTLayer')
-class CTLayer(tf.keras.layers.Layer):
-    def call(self, inputs):
-        # inputs shape: (batch, 2)
-        ups = tf.sigmoid(inputs[:, 0:1])  # keepdims -> (batch, 1)
-        X_C = relu_plus_one(inputs[:, 1:2])  # (batch, 1)
-        return tf.concat([ups, X_C], axis=-1)  # (batch, 2)
-
-    def compute_output_shape(self, input_shape):
-        # input_shape is (batch, 2) -> output_shape is (batch, 2)
-        return input_shape[:-1] + (2,)
-
-    def get_config(self):
-        # If there are no special args, only return super() config
-        return super().get_config()
-
-    @classmethod
-    def from_config(cls, config):
-        return cls(**config)
-
-@tf.keras.utils.register_keras_serializable(package='bdeissct_dl', name='loss_ct')
-def loss_ct(y_true, y_pred):
-
-    # Unpack the true values
-    p_true = y_true[:, 0]
-    X_true = y_true[:, 1]
-
-    # Unpack the predicted values
-    p_pred = y_pred[:, 0]
-    X_pred = y_pred[:, 1]
-
-    # Relative error for X_C
-    X_loss = tf.abs((X_pred - X_true) / X_true)
-
-    # Absolute error for ups
-    p_loss = tf.abs(p_pred - p_true)
-    # p_loss = tf.abs((p_pred - p_true) / tf.maximum(p_true, 1e-2))
-
-    mask = tf.cast(tf.greater(p_true, 1e-6), tf.float32)
-    X_loss = tf.reduce_mean(mask * X_loss)
-
-    # Combine the losses
-    return tf.reduce_mean(X_loss + p_loss)
-
-@tf.keras.utils.register_keras_serializable(package='bdeissct_dl', name='loss_ss')
-def loss_ss(y_true, y_pred):
-
-    # Unpack the true values
-    p_true = y_true[:, 0]
-    X_true = y_true[:, 1]
-
-    # Unpack the predicted values
-    p_pred = y_pred[:, 0]
-    X_pred = y_pred[:, 1]
-
-    # Relative error for X_S
-    X_loss = tf.abs((X_pred - X_true) / X_true)
-
-    # Absolute error for f_S, multiplied by 2, as f_S is in [0, 0.5]
-    p_loss = 2 * tf.abs(p_pred - p_true)
-    # p_loss = tf.abs((p_pred - p_true) / tf.maximum(p_true, 1e-2 / 2))
-
-    mask = tf.cast(tf.greater(p_true, 1e-6), tf.float32)
-    X_loss = tf.reduce_mean(mask * X_loss)
-
-    # Combine the losses
-    return tf.reduce_mean(X_loss + p_loss)
-
-
-@tf.keras.utils.register_keras_serializable(package='bdeissct_dl', name='loss_prob')
-def loss_prob(y_true, y_pred):
-    return tf.reduce_mean(tf.abs((y_pred - y_true) / tf.maximum(y_true, 1e-2)))
-
 @register_keras_serializable(package="bdeissct_dl", name="half_sigmoid")
 def half_sigmoid(x):
     return 0.5 * tf.sigmoid(x)  # range ~ [0, 0.5)
@@ -130,16 +27,12 @@ def relu_plus_one(x):
 
 
 LOSS_FUNCTIONS = {
-    LA: "mean_absolute_percentage_error",
-    PSI: "mean_absolute_percentage_error",
     REPRODUCTIVE_NUMBER: "mean_absolute_percentage_error",
     INFECTION_DURATION: "mean_absolute_percentage_error",
-    UPS_X_C: loss_ct,
+    INCUBATION_PERIOD: "mean_absolute_percentage_error",
     UPSILON: 'mae',
     RHO: 'mean_absolute_percentage_error',
     X_C: "mean_absolute_percentage_error",
-    F_E: 'mae',
-    F_S_X_S: loss_ss,
     F_S: 'mae',
     X_S: "mean_absolute_percentage_error",
 }
@@ -165,7 +58,6 @@ def build_model(target_columns, n_x, optimizer=None, metrics=None):
     x = tf.keras.layers.Dense(64, activation='elu', name=f'layer2_dense128_elu')(x)
     x = tf.keras.layers.Dropout(0.5, name='dropout2_50')(x)
     x = tf.keras.layers.Dense(32, activation='elu', name=f'layer3_dense64elu')(x)
-    # x = tf.keras.layers.Dropout(0.5, name='dropout3_50')(x)
     x = tf.keras.layers.Dense(16, activation='elu', name=f'layer4_dense32_elu')(x)
 
     outputs = {}
@@ -174,16 +66,12 @@ def build_model(target_columns, n_x, optimizer=None, metrics=None):
         outputs[REPRODUCTIVE_NUMBER] = tf.keras.layers.Dense(1, activation="softplus", name=REPRODUCTIVE_NUMBER)(x) # positive values only
     if INFECTION_DURATION in target_columns:
         outputs[INFECTION_DURATION] = tf.keras.layers.Dense(1, activation="softplus", name=INFECTION_DURATION)(x) # positive values only
-    if F_E in target_columns:
-        outputs[F_E] = tf.keras.layers.Dense(1, activation="sigmoid", name=F_E)(x)
-    # if F_S in target_columns:
-    #     outputs[F_S_X_S] = SSLayer(name=F_S_X_S)(tf.keras.layers.Dense(2, activation=None, name="FS_XS_logits")(x))
+    if INCUBATION_PERIOD in target_columns:
+        outputs[INCUBATION_PERIOD] = tf.keras.layers.Dense(1, activation="softplus", name=INCUBATION_PERIOD)(x) # positive values only
     if F_S in target_columns:
         outputs[F_S] = tf.keras.layers.Dense(1, activation=half_sigmoid, name="FS_logits")(x)
     if X_S in target_columns:
         outputs[X_S] = tf.keras.layers.Dense(1, activation=relu_plus_one, name="XS_logits")(x)
-    # if UPSILON in target_columns:
-    #     outputs[UPS_X_C] = CTLayer(name=UPS_X_C)(tf.keras.layers.Dense(2, activation=None, name="ups_XC_logits")(x))
     if UPSILON in target_columns:
         outputs[UPSILON] = tf.keras.layers.Dense(1, activation="sigmoid", name="ups_logits")(x)
     if X_C in target_columns: