darwin.js - crossover/trendline strategy input & general .csv/.json output issues.

[shr00mie]

## darwin.js - Crossover & Trendline - parameter inconsistencies:

Existing strategies are using markup_pct:

Newly added crossover and trendline instead utilize markdown_buy_pct & markup_sell_pct.

(btw, should these starting ranges extend into the negative?)

Results in undefined output in .csv/.json due to:

Since conf.js is utilizing the syntax from crossover/trendline, I suggest that darwin.js & phenotype.js be adjusted accordingly.

While we're at it, let's toss the ema/rsi section up top and move these back into the individual strategies for aesthetic and consistency purposes.

## darwin.js - .csv & .json output - issues due due to inconsistent variable input/parse.
Some previous modifications to how the selector data is pulled/stored and presented (krystoph mentioned "objectify selector PR") and the recent addition of --start --end have rather decently f'd up .csv and .json output:

CSV output issues originate here:

if selector on 628 is modified to selector.normalized, then column A in the .csv subsequently displays correctly as the normalized selector instead of the JSON output. issue with days probably has to do input being defined here:

(why is start colored different? conflict?)

instead of here:

i THINK the .csv/.json output for the sim params component is initially being being stored here:

suspiciously, selector is missing from both sections. so i'm not sure if it's being properly defined as selector.normalized once pulled or what it's being set as through the process to be output as json instead of what we're expecting.

there's this portion of code which i think is trying to pull the params output and normalize the selector, but if the selector is not being properly set or is not formatted as expected, this would not have the intended outcome.

[shr00mie]

both of these have now been remedied. closing.