update readme

README.md

@@ -1,6 +1,9 @@
-# Hawkes-ADR
+# Hawkes Process Modeling of Adverse Drug Reactions
+---
 
-A C++ implementation of our paper, "Hawkes Process Modeling of Adverse Drug Reactions with Longitudinal Observational Data"
+## About
+
+This repo contains a C++ implementation of our paper, "Hawkes Process Modeling of Adverse Drug Reactions with Longitudinal Observational Data" [[PDF]](http:https://people.csail.mit.edu/yujia/assets/pdf/hawkes_pdf.pdf).
 
 > @inproceedings{hawkes, 
 >  title={Hawkes Process Modeling of Adverse Drug Reactions with Longitudinal Observational Data}, 
@@ -9,18 +12,20 @@ A C++ implementation of our paper, "Hawkes Process Modeling of Adverse Drug Reac
 >  year={2017}
 > }
 
-## Data format
+## Data
 Each row in the data file represent the trajectory of one patient, which is a sequence of ordered time-event pairs separated by whitespace:
 > t_1 m_1 t_2 m_2 ... t_n m_n
 
-where $0\leq t_1\leq\cdots\leq t_n$ and $m_i\in\{1,\ldots,\text{numOfVariables}\}$. We assume $\{1,\ldots,\text{numOfOutcomes}\}$ are the indices for the adverse outcomes and $\{\text{numOfOutcomes}+1,\ldots,\text{numOfVariables}\}$ are the idices for the drugs of interest.
-
-## 
-
-## Todo:
+where 0 <= t_1 <= ... <= t_n and m_i in {1,...,numOfVariables}. We assume {1,...,numOfOutcomes} to be the indices for the adverse outcomes and {numOfOutcomes+1,...,numOfVariables} to be the idices for the drugs of interest.
 
-=======
+The Marshfield Clinic EHR data is not publicly available due to patients' privacy. We provide an example synthetic data generated from a Poisson autoregressive model. 
 
-## Todo:
-+ Data format
-+ Running on the synthetic example
+## Code Usage
+1. Compile
+```
+g++ --std=c++14 Hawkes.cpp -o Hawkes.out
+```
+2. Example run
+```
+./Hawkes.out -f "file_to_data" -k num_of_kernels -w length_of_windows -l lasso_regularizations 
+```