Fixed MD formatting for 2 tasks

ec2/analysis-tasks/az-tilt-estimation/assets/SubmissionInstructions.md

@@ -1,9 +1,8 @@
 # Submission Instructions
 
-### Validation Data Structure
+## Validation Data Structure
 
-Each system can have multiple inverter-level AC power data streams that can be analyzed, with each data stream validated independently of all other data streams in the system. 
-Each data stream is a Pandas series, with a timezone-aware Pandas datetime index, and an AC power time series in kW. 
+Each system can have multiple inverter-level AC power data streams that can be analyzed, with each data stream validated independently of all other data streams in the system. Each data stream is a Pandas series, with a timezone-aware Pandas datetime index, and an AC power time series in kW.
 
 An example time series is shown in Figure 1. An example dataset is provided for reference here.
 
@@ -16,6 +15,7 @@ Download an example data set here.
 ### File Submission Requirements
 
 The PV Validation Hub only accepts .zip file submissions. Each zip file must containing the following files:
+
 1) **submission_wrapper.py**: This Python file contains the master function for running the routines. Please refer to **Submitted Function Structure** below for more information on the associated function structure.
 2) **requirements.txt**: This text file contains the associated python packages (and versions, if applicable) to pip-install in order to run the submission_wrapper.py file.
 3) If applicable, any additional .py files containing support functions for **submission_wrapper.py**
@@ -24,7 +24,7 @@ The PV Validation Hub only accepts .zip file submissions. Each zip file must con
 
 In the **submission_wrapper.py** file, a standard function name must be used when validating a submitted algorithm. The standard function definition, which is required to run the validation, is as follows:
 
-```
+```python
 def estimate_az_tilt(time_series, latitude=None, longitude=None, data_sampling_frequency=None):
     """
     This is the required function definition for running azimuth-tilt estimation.
@@ -57,7 +57,7 @@ def estimate_az_tilt(time_series, latitude=None, longitude=None, data_sampling_f
 An automated results report containing the following metrics is returned following successful analysis:
 
 1) Mean average error (MAE) across all of the test cases, for both azimuth and tilt values. To calculate MAE, the ground-truth azimuth/tilt value is compared to its associated estimated value, and error is returned in degrees.
-3) Associated graphics, including error distributions for azimuth, tilt, and run time. Additionally, a scatter plot comparing algorithm run time vs. number of days in the AC power data set is provided. Example graphics are shown in Figures 3-6, respectively.
+2) Associated graphics, including error distributions for azimuth, tilt, and run time. Additionally, a scatter plot comparing algorithm run time vs. number of days in the AC power data set is provided. Example graphics are shown in Figures 3-6, respectively.
 
 ![image info](./absolute_error_az_dist.png)
 
@@ -74,5 +74,3 @@ An automated results report containing the following metrics is returned followi
 ![image info](./run_time_number_days.png)
 
 *Figure 6: Algorithm run time vs. number of days present in the AC power data tested.*
-
-

ec2/analysis-tasks/az-tilt-estimation/assets/description.md

@@ -1,7 +1,6 @@
 # Description
-It is not uncommon for PV installation metadata to be improperly labeled, including azimuth, tilt, and mounting configuration. Errors and missing data can be especially frequent in solar fleets that have been recently 
-acquired. Incorrect values for this critical metadata can affect the overall performance of PV system performance modelling. 
+
+It is not uncommon for PV installation metadata to be improperly labeled, including azimuth, tilt, and mounting configuration. Errors and missing data can be especially frequent in solar fleets that have been recently acquired. Incorrect values for this critical metadata can affect the overall performance of PV system performance modelling.
 
 This analysis benchmarks algorithm performance for correctly identifying azimuth and tilt for fixed tilt systems using AC power time series data as an input. Real-world AC power data from fixed-tilt systems is used to assess algorithm performance.
 These systems have been manually reviewed to ensure correct azimuth and tilt values. For each system, ground truth azimuth and tilt values are compared to algorithm-estimated azimuth and tilt values and error in degrees is calculated.
-

ec2/analysis-tasks/az-tilt-estimation/assets/shortdesc.md

@@ -1 +1 @@
-Estimate azimuth and tilt values for fixed-tilt PV systems.
+# Estimate azimuth and tilt values for fixed-tilt PV systems.

ec2/analysis-tasks/time-shift-detection/assets/SubmissionInstructions.md

@@ -1,8 +1,9 @@
 # Submission Instructions
 
-### Validation Data Structure
+## Validation Data Structure
 
 For each AC power stream analyzed, there are two sets of data:
+
 1) The associated time series data for the AC power, in kW, and an index for timezone-aware datetime.
 2) The time shift validation file, which gives the associated time shift amount on a daily basis. A daily timezone-unaware timestamp is used as the associated index. 
 
@@ -21,6 +22,7 @@ Download an example data set here.
 ### File Submission Requirements
 
 The PV Validation Hub only accepts .zip file submissions. Each zip file must containing the following files:
+
 1) **submission_wrapper.py**: This Python file contains the master function for running the routines. Please refer to **Submitted Function Structure** below for more information on the associated function structure.
 2) **requirements.txt**: This text file contains the associated python packages (and versions, if applicable) to pip-install in order to run the submission_wrapper.py file.
 3) If applicable, any additional .py files containing support functions for **submission_wrapper.py**
@@ -29,7 +31,7 @@ The PV Validation Hub only accepts .zip file submissions. Each zip file must con
 
 In the **submission_wrapper.py** file, a standard function name must be used when validating a submitted algorithm. The standard function definition, which is required to run the validation, is as follows:
 
-```
+```python
 def detect_time_shifts(time_series,
                        latitude=None, longitude=None,
                        data_sampling_frequency=None):

ec2/analysis-tasks/time-shift-detection/assets/description.md

@@ -1,7 +1,9 @@
 
 # Description
+
 Measured photovolataic (PV) data from fielded systems may be incorrectly labeled by time zone, erroneously
-contain daylight savings time (DST), or have time shift issues due to operator error. Correcting time shifts in PV data is essential for accurate solar performance analysis, including degradation analysis. To illustrate this associated issue, Figures 1, 2, and 3, respectively, show AC power data streams heatmapped by hour of the day for the following scenarios: 
+contain daylight savings time (DST), or have time shift issues due to operator error. Correcting time shifts in PV data is essential for accurate solar performance analysis, including degradation analysis. To illustrate this associated issue, Figures 1, 2, and 3, respectively, show AC power data streams heatmapped by hour of the day for the following scenarios:
+
 1) No issues
 2) DST is present
 3) A random time shift is present
@@ -17,5 +19,4 @@ contain daylight savings time (DST), or have time shift issues due to operator e
 
 Ideally, we want to run analysis on an AC power data stream like Figure 1, where no random time shifts or DST is present.
 
-This analysis benchmarks algorithm performance for finding time shifts in PV data and correcting them accordingly. To do this, real-world PV data sets are manipulated to create common time shift issues, such as incorrect time zone, partial DST, full DST, and random time shifts. All ground truth time shift issues for each data set are recorded, and used here to validate submitted time shift detection algorithms. In particular, the associated time shift amount at each value in the predicted time series will be compared to the associated ground-truth time shift value. 
-
+This analysis benchmarks algorithm performance for finding time shifts in PV data and correcting them accordingly. To do this, real-world PV data sets are manipulated to create common time shift issues, such as incorrect time zone, partial DST, full DST, and random time shifts. All ground truth time shift issues for each data set are recorded, and used here to validate submitted time shift detection algorithms. In particular, the associated time shift amount at each value in the predicted time series will be compared to the associated ground-truth time shift value.