Robert’s Working Blog

1. programming for mean average trajectory to test the validility

2. calculate the biggest differnce between term and preterm

3. working on difference plot for each time spot 

1. Calculate the difference between “Term” and “Preterm” at the same time to plot a difference graph. This can predict what time is the biggest difference; To identify the key time point;

2. Identify the difference patterms between “Term” and “Preterm” for CRH, hCG, Prog and Cortisol.

3. Weekly look at the plot to identify the difference 

1. Working for Bruce to make dynamic images

2. Calculating SD and find problem on mean value (I need to re-calculate the value for different time) 

1. Regression line (regression method)

2. Correlation coefficient

3.  RMS Error (Regression Measurement Error) 

1. Root Mean Square (Mean, Median, SD)

2. Coefficient 

1. Reading Stastics books about “Sensitivity, Specification and ROC curve”

http://gim.unmc.edu/dxtests/howto.htm

Meeting with David again:

1. Need to pay attention to “prog”

2. May plot based on different weeks (37-38; 38-39; 39-40) 

Meeting with David and need to:

1. create mean/average trajectory

2. cut trajectory based on differnt time and calculate the differnt CI

3. counting the crossing section for each trajectory

4. filter data to define relationship(parallel and crossing)

5. Need to look at “Specificity”, “Sensitivity” and “ROC curves” 

1. Working on plotting single parameter with time

2. Confirm the correctness of the plotting.

3. Start to build three plotting together (CRH, hCG and Cortisol).

1. Plotting individual contour for both term and pre-term

2. make dynamic images for this

3. modify group website

4. working on plotting individual parameter with time